Optical flow evaluation results Statistics:     Average   SD   R0.5   R1.0   R2.0   A90   A95   A99  
Error type:   endpoint   angle   interpolation   normalized interpolation  
Show images: below table   above table        
R0.5
normalized interpolation
error		 avg. Mequon
(Hidden texture)
im0   GT   im1 		Schefflera
(Hidden texture)
im0   GT   im1 		Urban
(Synthetic)
im0   GT   im1 		Teddy
(Stereo)
im0   GT   im1 		Backyard
(High-speed camera)
im0   GT   im1 		Basketball
(High-speed camera)
im0   GT   im1 		Dumptruck
(High-speed camera)
im0   GT   im1 		Evergreen
(High-speed camera)
im0   GT   im1
rank all disc untext all disc untext all disc untext all disc untext all disc untext all disc untext all disc untext all disc untext
SoftSplat [169]5.9 32.8 14 24.6 7 46.3 16 20.8 3 24.0 4 34.4 3 18.1 1 17.9 2 35.0 1 47.7 2 38.9 2 76.9 2 64.5 16 57.5 13 82.5 6 63.1 4 39.8 6 79.7 3 26.7 7 36.1 11 41.5 8 29.2 3 39.9 5 48.3 3
IFRNet [193]8.8 33.6 16 24.2 6 47.5 18 21.9 6 23.9 3 36.5 6 18.7 4 18.2 3 35.7 4 48.7 6 39.8 4 78.7 16 63.9 7 56.8 6 83.6 18 64.3 9 42.2 9 80.5 17 26.5 6 35.3 10 41.8 10 31.3 9 44.0 8 49.2 10
EAFI [186]9.8 30.5 8 22.0 2 43.5 12 19.3 1 19.8 1 32.0 1 18.2 2 17.2 1 35.0 1 47.2 1 38.1 1 76.3 1 67.8 26 62.6 26 82.1 1 65.4 21 46.1 22 79.9 6 27.7 15 37.7 16 41.9 12 33.3 12 47.6 18 49.9 28
SoftsplatAug [190]10.3 28.8 6 22.0 2 40.9 7 19.4 2 21.9 2 32.9 2 18.6 3 19.2 4 35.4 3 48.5 4 40.0 5 77.6 6 62.1 2 54.5 2 82.3 5 62.7 2 38.5 3 79.7 3 27.4 14 35.1 8 42.7 149 29.2 3 39.3 4 48.8 6
DistillNet [184]11.6 32.9 15 25.4 10 46.2 15 21.7 4 24.8 5 35.6 5 19.3 5 19.7 6 36.6 5 48.3 3 40.1 6 77.2 3 64.4 14 57.7 17 82.2 4 64.9 19 45.2 21 79.9 6 28.0 20 40.2 24 41.8 10 34.0 20 49.0 21 49.8 21
SepConv++ [185]17.5 35.3 21 28.7 24 48.6 126 24.5 9 28.4 13 39.4 125 21.4 9 22.7 9 36.8 6 51.2 18 43.5 18 78.6 14 63.0 4 55.5 3 82.8 10 61.6 1 35.2 1 79.3 1 25.7 1 32.9 2 41.3 3 26.8 1 35.2 1 47.9 1
MV_VFI [183]24.2 35.6 22 28.3 21 49.4 148 25.8 21 30.4 28 39.7 129 21.3 8 23.1 12 38.0 9 50.2 10 43.0 13 78.1 10 64.4 14 57.6 15 83.1 12 64.3 9 43.5 16 79.9 6 27.1 10 38.9 17 41.4 4 33.7 15 49.4 23 49.1 9
TC-GAN [166]24.8 35.6 22 28.6 23 49.6 154 25.8 21 30.6 34 39.6 128 21.2 7 23.0 11 37.9 8 50.3 13 43.1 14 78.3 12 64.3 12 57.5 13 83.0 11 64.3 9 43.4 13 79.9 6 27.2 13 39.0 19 41.4 4 33.7 15 49.3 22 49.2 10
DAI [168]25.9 35.2 20 24.8 8 50.0 164 23.8 7 27.2 7 36.9 10 19.5 6 19.4 5 36.9 7 48.6 5 39.7 3 78.6 14 68.7 27 63.0 27 83.1 12 66.1 23 48.2 26 80.2 16 28.6 116 40.1 23 42.3 15 34.8 26 50.2 26 49.9 28
STSR [170]28.7 34.2 17 27.0 15 47.9 42 21.7 4 25.5 6 34.9 4 22.7 14 24.7 17 38.1 10 49.6 7 41.5 9 78.7 16 70.2 29 65.2 30 84.7 24 66.7 24 50.4 29 80.8 19 28.6 116 43.0 35 42.6 135 35.3 27 53.1 32 49.9 28
DAIN [152]31.0 36.5 103 28.9 27 50.8 176 26.2 45 30.6 34 40.0 132 21.4 9 23.3 13 38.1 10 50.2 10 43.1 14 77.9 7 64.5 16 57.6 15 83.4 16 64.3 9 43.5 16 80.0 11 27.1 10 39.1 20 41.4 4 33.7 15 49.4 23 49.2 10
AdaCoF [165]31.1 37.1 142 29.5 30 50.7 175 26.1 33 29.6 20 40.6 142 24.6 16 24.4 15 38.5 17 51.3 19 43.3 17 78.0 8 67.5 25 60.9 24 85.0 26 62.9 3 39.0 5 79.6 2 26.1 4 35.1 8 41.2 1 29.5 6 41.3 7 48.2 2
GDCN [172]35.5 31.1 9 28.3 21 41.8 8 29.3 153 33.3 103 41.4 153 21.8 12 22.9 10 39.3 107 52.6 87 44.2 22 78.5 13 64.3 12 57.2 11 84.0 21 64.7 16 43.4 13 80.6 18 26.9 9 36.2 13 42.2 14 31.1 8 45.1 11 49.0 8
IDIAL [192]37.4 31.6 11 25.9 11 43.3 11 25.3 13 28.2 11 37.8 96 21.5 11 21.7 8 38.9 52 50.2 10 42.3 11 77.5 5 63.7 5 56.9 7 82.7 9 64.8 17 44.5 19 79.9 6 29.5 171 38.9 17 43.8 182 34.4 23 47.5 16 52.0 176
EDSC [173]38.0 31.7 12 26.4 13 43.6 13 24.5 9 28.5 14 40.4 137 24.6 16 23.9 14 44.4 176 51.0 15 43.5 18 78.7 16 63.9 7 56.7 5 83.3 15 64.6 15 42.3 10 80.9 20 26.8 8 34.2 4 43.4 173 32.4 11 44.3 10 53.1 182
MDP-Flow2 [68]38.1 35.7 24 30.6 35 47.8 31 25.9 25 30.5 31 36.9 10 28.6 29 29.8 32 38.5 17 51.9 29 46.5 48 80.3 33 71.9 38 66.6 36 87.2 43 68.6 36 53.9 61 82.1 66 28.1 25 43.6 46 42.4 32 36.6 72 55.6 76 50.0 40
BMBC [171]39.0 36.0 43 27.1 16 49.4 148 26.2 45 27.5 9 43.1 167 31.9 184 30.3 39 45.0 181 49.6 7 40.7 7 78.1 10 64.2 10 57.0 8 83.1 12 63.5 5 39.8 6 80.1 14 26.0 3 34.5 5 41.2 1 29.3 5 40.7 6 48.6 4
PMMST [112]39.5 35.8 30 30.8 36 47.9 42 26.5 60 31.0 45 37.3 53 28.6 29 29.9 33 38.4 12 51.7 23 46.0 34 80.2 26 72.0 43 66.7 42 87.3 52 68.5 33 53.3 39 82.0 38 28.1 25 43.7 52 42.4 32 36.5 60 55.5 70 50.0 40
PH-Flow [99]40.1 36.1 51 32.5 65 47.8 31 25.6 16 29.6 20 36.9 10 28.7 37 30.0 35 38.5 17 51.6 21 45.5 29 80.2 26 71.9 38 66.7 42 87.3 52 68.8 74 54.8 118 81.9 28 28.1 25 43.6 46 42.4 32 36.4 52 55.3 57 50.0 40
NNF-Local [75]40.3 35.7 24 31.4 39 47.6 20 25.5 14 29.6 20 36.9 10 28.6 29 29.9 33 38.5 17 52.4 63 48.0 104 80.3 33 72.0 43 66.6 36 87.4 78 68.7 52 54.4 91 82.0 38 28.1 25 43.5 43 42.4 32 36.2 38 55.0 46 50.0 40
CoT-AMFlow [174]40.6 35.7 24 30.5 34 47.7 27 26.0 28 30.6 34 37.0 20 28.7 37 30.4 44 38.5 17 51.8 25 46.2 39 80.3 33 72.1 58 66.7 42 87.3 52 68.7 52 54.0 66 82.1 66 28.1 25 43.5 43 42.4 32 36.5 60 55.6 76 50.0 40
MEMC-Net+ [160]41.8 36.9 136 28.9 27 50.5 171 27.0 87 29.9 24 41.3 150 23.1 15 24.6 16 39.3 107 51.1 16 42.7 12 78.0 8 67.3 24 61.2 25 83.8 19 65.6 22 47.9 24 80.1 14 27.7 15 40.7 25 41.7 9 33.9 19 50.6 27 49.2 10
STAR-Net [164]44.0 36.8 130 26.4 13 51.8 182 26.2 45 28.3 12 40.0 132 22.3 13 20.8 7 39.1 81 49.9 9 41.6 10 77.2 3 62.9 3 55.7 4 82.1 1 64.4 13 43.5 16 80.0 11 29.0 154 36.1 11 42.3 15 33.8 18 46.7 14 50.8 160
NN-field [71]46.8 36.0 43 32.2 55 47.9 42 25.5 14 29.3 18 36.8 8 29.4 99 29.7 31 39.0 71 52.4 63 48.1 111 80.3 33 72.0 43 66.7 42 87.3 52 68.7 52 54.0 66 82.0 38 28.1 25 43.4 40 42.4 32 36.4 52 55.2 53 50.0 40
MS_RAFT+_RVC [195]48.6 36.2 59 32.3 58 48.2 93 26.1 33 31.7 62 37.0 20 28.5 27 29.4 28 38.4 12 51.8 25 46.3 41 80.3 33 72.3 111 67.0 82 87.4 78 68.7 52 53.7 48 82.1 66 27.8 17 41.6 26 42.4 32 36.3 41 56.1 107 49.5 15
ProbFlowFields [126]51.9 35.9 36 32.4 60 48.0 59 25.8 21 30.5 31 37.2 46 28.6 29 30.3 39 38.5 17 52.1 43 46.4 44 80.7 100 72.3 111 67.1 114 87.5 140 68.6 36 53.8 51 82.1 66 28.0 20 42.8 30 42.3 15 36.1 34 54.6 40 50.1 63
ADC [161]52.0 37.5 153 30.3 32 50.5 171 28.6 137 31.3 54 44.9 177 26.3 20 27.0 25 39.6 131 53.6 149 45.6 30 79.4 19 66.5 22 59.9 22 84.2 23 64.1 8 42.9 12 80.0 11 26.1 4 34.9 6 41.4 4 33.3 12 48.9 20 48.9 7
IROF++ [58]52.6 36.2 59 33.0 85 47.8 31 26.1 33 30.9 42 36.9 10 29.1 65 31.0 66 38.9 52 51.6 21 45.6 30 80.4 48 72.0 43 66.8 54 87.2 43 68.6 36 53.4 40 82.2 97 28.3 48 44.6 89 42.4 32 36.5 60 55.3 57 50.4 121
FGME [158]52.8 24.9 1 21.5 1 33.9 1 26.1 33 28.0 10 43.1 167 27.2 24 24.8 18 48.5 187 50.6 14 41.0 8 80.8 116 61.7 1 54.0 1 82.5 6 64.4 13 38.8 4 82.2 97 29.6 175 31.2 1 54.2 190 29.9 7 38.3 3 55.3 189
DSepConv [162]52.8 34.2 17 28.8 26 46.6 17 27.8 117 31.4 56 43.5 171 26.1 18 25.1 19 44.5 177 52.7 98 44.9 25 79.7 20 64.2 10 57.1 9 84.0 21 65.0 20 42.8 11 81.2 22 27.1 10 35.0 7 43.5 175 34.0 20 47.5 16 53.6 185
Sparse-NonSparse [56]54.6 36.2 59 32.8 78 48.0 59 25.9 25 30.4 28 37.0 20 29.0 55 30.9 62 38.8 35 52.0 35 46.1 36 80.6 75 72.1 58 66.8 54 87.3 52 68.9 88 54.6 105 82.1 66 28.3 48 44.0 62 42.4 32 36.4 52 55.4 64 50.1 63
CombBMOF [111]55.9 35.9 36 31.0 37 47.8 31 25.8 21 30.5 31 36.8 8 29.2 73 30.8 57 39.5 120 52.4 63 47.4 76 80.3 33 72.1 58 66.8 54 87.4 78 68.9 88 54.5 98 82.1 66 28.5 97 44.6 89 42.3 15 36.0 33 54.6 40 50.0 40
nLayers [57]58.2 36.4 91 32.0 50 48.2 93 26.0 28 30.4 28 37.3 53 28.7 37 29.4 28 38.8 35 52.2 51 46.8 55 80.4 48 72.3 111 67.1 114 87.4 78 68.8 74 54.7 110 82.0 38 28.3 48 43.7 52 42.4 32 36.4 52 55.4 64 49.9 28
AGIF+OF [84]58.5 36.2 59 32.8 78 47.9 42 26.1 33 30.8 39 37.1 31 29.0 55 30.7 53 38.9 52 51.8 25 46.2 39 80.1 24 72.3 111 67.2 134 87.3 52 68.9 88 55.2 137 81.9 28 28.3 48 43.6 46 42.4 32 36.6 72 56.0 98 49.9 28
GMFlow_RVC [196]59.5 36.0 43 32.6 70 47.9 42 26.1 33 31.7 62 37.2 46 28.7 37 30.6 48 38.5 17 52.3 56 47.8 94 80.3 33 72.4 145 67.1 114 87.3 52 68.8 74 54.7 110 82.0 38 28.2 36 43.4 40 42.5 95 36.4 52 55.5 70 49.8 21
2DHMM-SAS [90]60.9 36.4 91 33.9 127 47.9 42 27.1 92 32.6 83 37.0 20 28.5 27 30.4 44 38.9 52 51.8 25 45.6 30 80.4 48 72.1 58 66.9 67 87.4 78 68.8 74 54.5 98 82.0 38 28.3 48 44.2 71 42.3 15 36.7 87 56.1 107 50.0 40
NNF-EAC [101]62.0 36.3 78 32.4 60 48.0 59 26.6 65 31.7 62 37.1 31 29.3 84 30.2 36 39.0 71 52.4 63 46.9 58 81.1 150 72.0 43 66.7 42 87.4 78 68.7 52 53.7 48 82.1 66 28.2 36 43.9 56 42.4 32 36.7 87 55.9 92 50.0 40
Layers++ [37]62.7 36.3 78 32.4 60 48.2 93 25.7 17 29.2 17 37.3 53 28.9 51 30.6 48 38.9 52 52.0 35 46.4 44 80.4 48 72.2 76 67.0 82 87.5 140 68.9 88 55.2 137 82.0 38 28.3 48 44.0 62 42.4 32 36.6 72 55.5 70 50.1 63
RAFT-TF_RVC [179]62.9 35.8 30 32.6 70 47.6 20 26.1 33 31.4 56 37.1 31 28.7 37 30.8 57 38.5 17 52.4 63 48.4 130 80.4 48 72.3 111 67.1 114 87.3 52 70.6 189 54.4 91 83.9 187 28.0 20 42.9 33 42.4 32 35.8 32 54.4 37 49.7 19
PRAFlow_RVC [177]63.6 36.1 51 31.9 46 47.9 42 26.3 53 31.6 60 37.4 65 28.6 29 30.3 39 38.4 12 52.5 71 48.0 104 80.5 67 72.2 76 66.9 67 87.4 78 68.6 36 53.4 40 82.2 97 28.1 25 43.1 36 42.5 95 37.2 143 56.5 131 50.1 63
LSM [39]64.0 36.3 78 33.7 116 48.0 59 26.1 33 31.0 45 37.0 20 29.1 65 31.8 86 38.9 52 52.2 51 46.9 58 80.6 75 72.1 58 66.9 67 87.3 52 69.0 104 54.9 122 82.1 66 28.3 48 44.1 68 42.4 32 36.5 60 55.7 81 50.0 40
ComponentFusion [94]64.2 36.0 43 32.2 55 48.0 59 26.1 33 31.1 50 36.9 10 29.1 65 32.3 96 38.8 35 52.0 35 47.0 62 80.3 33 72.2 76 67.1 114 87.3 52 68.7 52 53.9 61 82.1 66 28.5 97 46.1 152 42.4 32 36.7 87 55.8 89 50.2 87
RAFT-it+_RVC [198]64.5 35.7 24 31.9 46 47.6 20 26.0 28 31.4 56 36.9 10 28.8 43 31.5 77 38.4 12 52.4 63 48.3 124 80.2 26 72.3 111 67.1 114 87.4 78 69.8 164 55.5 152 83.2 174 28.0 20 42.8 30 42.5 95 35.5 29 53.7 33 49.7 19
VCN_RVC [178]64.8 36.2 59 33.5 104 47.9 42 26.4 56 32.0 67 37.2 46 29.6 113 34.7 156 38.8 35 52.5 71 48.6 136 80.7 100 72.2 76 67.0 82 87.3 52 68.7 52 54.2 81 81.9 28 28.2 36 44.0 62 42.3 15 35.6 31 53.9 35 49.8 21
ProBoost-Net [191]65.7 25.7 3 23.5 4 34.6 3 26.8 73 29.9 24 41.8 160 26.9 22 25.4 20 48.0 185 52.2 51 43.8 20 82.2 175 66.9 23 60.5 23 85.2 27 67.0 27 46.5 23 83.0 170 28.6 116 37.4 15 45.9 187 34.4 23 47.4 15 54.9 188
FlowFields [108]65.8 36.0 43 32.7 75 47.9 42 26.4 56 32.0 67 37.3 53 29.0 55 32.6 105 38.7 28 52.5 71 47.9 99 80.7 100 72.3 111 67.0 82 87.5 140 68.6 36 54.4 91 82.0 38 28.2 36 44.0 62 42.4 32 36.3 41 55.2 53 50.1 63
CyclicGen [149]66.1 39.1 177 29.0 29 53.8 187 30.3 169 29.4 19 54.5 195 29.3 84 30.9 62 45.9 183 53.7 154 45.1 26 82.2 175 65.9 21 58.2 20 85.2 27 63.7 6 37.0 2 81.7 23 25.9 2 33.2 3 42.0 13 27.0 2 35.7 2 48.7 5
RAFT-it [194]66.1 35.7 24 31.9 46 47.6 20 25.7 17 30.6 34 36.7 7 28.6 29 30.5 47 38.4 12 52.3 56 47.9 99 80.2 26 72.3 111 67.0 82 87.4 78 70.3 185 54.6 105 83.9 187 28.0 20 42.7 29 42.4 32 37.3 150 57.7 173 49.6 17
S2F-IF [121]66.3 35.9 36 32.5 65 47.8 31 26.2 45 31.6 60 37.2 46 29.0 55 31.9 91 38.6 26 52.3 56 47.6 81 80.4 48 72.4 145 67.2 134 87.5 140 68.7 52 54.5 98 81.9 28 28.4 71 44.7 94 42.4 32 36.3 41 55.2 53 50.1 63
TV-L1-MCT [64]66.4 36.8 130 34.7 153 48.2 93 26.7 67 32.4 80 37.3 53 28.6 29 30.9 62 39.0 71 51.9 29 45.7 33 80.5 67 72.2 76 67.0 82 87.3 52 68.6 36 53.0 35 82.3 122 28.3 48 44.4 78 42.4 32 36.1 34 54.9 44 50.2 87
FlowFields+ [128]66.5 35.9 36 32.6 70 47.9 42 26.4 56 32.2 73 37.4 65 29.0 55 32.6 105 38.7 28 52.3 56 47.7 87 80.6 75 72.3 111 67.1 114 87.5 140 68.7 52 54.6 105 82.0 38 28.2 36 44.0 62 42.4 32 36.3 41 55.2 53 50.1 63
FeFlow [167]66.5 29.0 7 26.1 12 39.5 6 28.1 123 31.2 51 45.4 180 27.1 23 26.3 24 49.8 193 52.0 35 44.3 23 80.6 75 63.9 7 57.2 11 82.6 8 66.9 26 48.0 25 82.3 122 33.9 192 39.1 20 61.7 193 35.5 29 48.6 19 56.4 191
WLIF-Flow [91]67.6 36.1 51 32.5 65 47.8 31 26.3 53 31.2 51 37.1 31 29.1 65 30.7 53 39.1 81 52.0 35 46.4 44 80.6 75 72.1 58 66.8 54 87.4 78 69.0 104 54.9 122 82.3 122 28.3 48 43.9 56 42.5 95 36.8 96 55.9 92 50.1 63
MPRN [151]68.3 36.2 59 28.7 24 49.4 148 29.4 155 32.0 67 43.3 169 32.4 186 37.1 182 42.2 170 52.1 43 45.2 28 80.0 21 70.8 30 65.1 29 86.7 33 67.5 28 49.4 27 82.2 97 27.9 18 42.3 28 42.3 15 34.5 25 51.6 29 49.9 28
HCFN [157]68.9 35.7 24 31.7 42 47.6 20 26.6 65 32.8 88 37.0 20 29.0 55 32.4 99 38.7 28 52.2 51 47.3 68 80.6 75 72.1 58 66.8 54 87.4 78 69.8 164 53.8 51 83.5 182 28.4 71 44.7 94 42.5 95 36.4 52 55.3 57 50.1 63
COFM [59]69.7 36.1 51 32.0 50 48.1 75 26.1 33 30.8 39 37.1 31 28.8 43 30.3 39 38.8 35 51.7 23 46.0 34 80.0 21 72.2 76 67.2 134 87.2 43 68.9 88 56.1 171 81.7 23 28.1 25 42.8 30 43.1 168 37.1 134 56.9 149 50.7 158
LME [70]70.8 35.8 30 31.0 37 47.8 31 26.9 80 32.2 73 38.4 113 29.2 73 32.6 105 38.8 35 51.9 29 46.7 53 80.4 48 72.6 174 67.4 162 87.7 185 68.8 74 54.9 122 82.0 38 28.1 25 43.5 43 42.4 32 36.3 41 55.3 57 50.0 40
FMOF [92]70.9 36.5 103 33.7 116 48.2 93 25.9 25 30.3 26 37.1 31 29.3 84 30.7 53 39.0 71 52.5 71 47.5 77 80.2 26 72.2 76 67.0 82 87.5 140 69.0 104 55.1 132 82.0 38 28.1 25 43.4 40 42.4 32 36.8 96 56.0 98 50.1 63
OFLAF [78]71.0 35.8 30 31.5 40 47.8 31 25.7 17 29.8 23 37.0 20 29.0 55 31.2 69 38.7 28 52.0 35 46.8 55 80.1 24 72.4 145 67.3 150 87.4 78 68.9 88 55.3 142 82.0 38 28.6 116 45.4 133 42.4 32 37.1 134 57.1 158 50.1 63
RNLOD-Flow [119]71.7 36.3 78 33.5 104 48.0 59 26.8 73 32.6 83 37.1 31 29.2 73 31.8 86 38.8 35 52.1 43 46.9 58 80.2 26 72.2 76 67.0 82 87.3 52 69.0 104 55.2 137 82.1 66 28.3 48 44.2 71 42.4 32 37.1 134 56.9 149 49.8 21
DeepFlow2 [106]72.5 36.2 59 32.4 60 48.2 93 27.1 92 32.9 91 37.8 96 29.2 73 32.9 112 39.0 71 52.5 71 47.5 77 80.5 67 72.2 76 66.9 67 87.5 140 68.5 33 52.9 34 82.1 66 28.3 48 44.4 78 42.4 32 36.4 52 55.4 64 50.2 87
EAI-Flow [147]73.3 36.3 78 32.5 65 48.2 93 27.2 96 33.2 99 38.4 113 29.4 99 33.1 121 39.0 71 52.2 51 47.3 68 80.3 33 72.2 76 67.1 114 87.3 52 68.7 52 53.5 44 82.1 66 28.4 71 44.8 103 42.4 32 36.2 38 54.5 38 50.2 87
FF++_ROB [141]73.7 36.0 43 32.6 70 47.9 42 26.8 73 32.6 83 37.6 85 29.3 84 32.4 99 38.9 52 52.5 71 48.3 124 80.5 67 72.4 145 67.1 114 87.4 78 68.8 74 54.3 86 82.2 97 28.2 36 44.0 62 42.4 32 36.3 41 55.1 48 50.1 63
MAF-net [163]73.7 25.1 2 23.6 5 34.1 2 25.2 12 29.1 15 41.4 153 26.1 18 26.2 23 48.6 188 52.7 98 44.7 24 82.6 179 69.4 28 63.7 28 85.4 29 68.1 29 49.9 28 83.7 186 37.6 193 39.6 22 78.4 197 36.7 87 50.8 28 60.1 195
DeepFlow [85]73.9 36.1 51 31.8 45 48.1 75 27.3 99 32.9 91 38.4 113 29.3 84 33.3 127 39.1 81 52.6 87 47.0 62 80.7 100 72.2 76 66.8 54 87.5 140 68.7 52 52.8 33 82.5 147 28.1 25 43.6 46 42.3 15 36.2 38 55.0 46 50.2 87
Ramp [62]74.0 36.5 103 34.0 132 48.2 93 26.0 28 30.8 39 37.1 31 28.9 51 30.8 57 38.8 35 51.9 29 46.1 36 80.4 48 72.2 76 67.0 82 87.4 78 69.1 118 55.4 148 82.2 97 28.4 71 44.7 94 42.4 32 36.8 96 56.2 116 50.2 87
MDP-Flow [26]75.2 35.8 30 31.5 40 48.0 59 26.2 45 31.4 56 37.4 65 29.0 55 31.1 67 38.9 52 52.7 98 47.8 94 80.7 100 72.2 76 66.9 67 87.5 140 68.9 88 55.2 137 82.1 66 28.5 97 45.3 132 42.5 95 36.3 41 55.4 64 50.0 40
IROF-TV [53]75.6 36.3 78 33.6 112 48.2 93 26.2 45 31.0 45 37.0 20 29.3 84 33.6 134 39.1 81 51.9 29 46.5 48 80.8 116 72.3 111 67.0 82 87.6 175 68.5 33 53.9 61 81.9 28 28.3 48 44.9 108 42.3 15 36.6 72 55.6 76 50.4 121
SegFlow [156]75.7 36.2 59 33.4 100 48.1 75 26.5 60 32.3 79 37.5 75 29.2 73 32.5 103 38.9 52 52.3 56 47.9 99 80.6 75 72.3 111 67.0 82 87.5 140 68.6 36 54.1 73 82.1 66 28.3 48 44.4 78 42.4 32 36.5 60 55.4 64 50.4 121
Classic+NL [31]76.5 36.5 103 34.0 132 48.2 93 26.2 45 30.9 42 37.1 31 28.8 43 30.6 48 38.8 35 52.1 43 46.5 48 80.6 75 72.2 76 67.0 82 87.4 78 69.2 131 55.3 142 82.2 97 28.4 71 44.6 89 42.4 32 36.8 96 56.2 116 50.2 87
PGM-C [118]76.9 36.2 59 33.3 95 48.1 75 26.5 60 32.2 73 37.5 75 29.2 73 32.9 112 38.8 35 52.5 71 48.3 124 80.7 100 72.3 111 67.0 82 87.5 140 68.6 36 54.0 66 82.0 38 28.3 48 44.6 89 42.4 32 36.5 60 55.5 70 50.4 121
DF-Auto [113]77.9 36.8 130 31.9 46 48.9 138 28.5 134 33.7 114 40.8 144 28.8 43 30.3 39 38.7 28 52.5 71 47.3 68 80.4 48 72.1 58 66.7 42 87.4 78 68.6 36 53.8 51 82.0 38 28.4 71 44.7 94 42.5 95 36.8 96 56.3 121 50.2 87
FC-2Layers-FF [74]79.1 36.4 91 33.8 122 48.1 75 25.7 17 29.1 15 37.4 65 28.9 51 30.9 62 38.8 35 52.1 43 46.8 55 80.6 75 72.3 111 67.2 134 87.4 78 69.1 118 55.5 152 82.1 66 28.4 71 44.7 94 42.5 95 36.9 113 56.3 121 50.0 40
FRUCnet [153]80.0 42.4 190 29.5 30 59.4 196 29.8 163 31.2 51 48.2 188 29.6 113 28.0 26 50.5 194 54.0 164 46.1 36 80.3 33 65.1 20 58.2 20 83.4 16 64.8 17 43.4 13 81.0 21 27.9 18 36.4 14 44.1 185 33.4 14 46.6 13 53.3 184
PBOFVI [189]80.5 36.7 123 35.0 158 48.1 75 27.6 109 33.8 117 37.5 75 29.6 113 31.2 69 39.2 97 52.3 56 47.7 87 80.3 33 72.3 111 67.1 114 87.4 78 68.8 74 54.2 81 82.1 66 28.4 71 44.3 74 42.4 32 36.1 34 54.9 44 50.0 40
HAST [107]80.7 36.1 51 31.7 42 48.1 75 26.1 33 31.0 45 37.0 20 29.3 84 31.7 81 39.2 97 51.9 29 46.5 48 80.3 33 72.3 111 67.3 150 87.2 43 69.3 143 56.4 179 82.0 38 28.4 71 45.0 116 42.5 95 37.3 150 57.3 162 50.0 40
CPM-Flow [114]83.1 36.2 59 33.5 104 48.1 75 26.5 60 32.2 73 37.5 75 29.3 84 32.6 105 38.9 52 52.7 98 48.7 143 80.7 100 72.3 111 67.0 82 87.5 140 68.7 52 53.8 51 82.2 97 28.4 71 44.7 94 42.4 32 36.5 60 55.5 70 50.3 107
Classic+CPF [82]83.4 36.4 91 33.6 112 47.9 42 26.3 53 31.3 54 37.0 20 28.8 43 31.1 67 38.9 52 52.0 35 46.5 48 80.0 21 72.5 161 67.4 162 87.4 78 69.2 131 56.1 171 82.0 38 28.6 116 45.2 129 42.4 32 37.2 143 57.3 162 50.0 40
Second-order prior [8]83.8 36.2 59 32.1 53 48.1 75 27.9 120 34.1 125 37.4 65 29.9 137 34.6 153 39.7 136 52.4 63 47.2 66 80.6 75 71.9 38 66.6 36 87.5 140 68.7 52 54.0 66 82.1 66 28.5 97 45.2 129 42.4 32 36.5 60 55.7 81 50.2 87
UnDAF [187]83.8 36.1 51 33.0 85 47.8 31 26.7 67 32.7 86 37.0 20 29.6 113 35.6 172 38.7 28 53.6 149 52.2 187 80.7 100 72.1 58 66.7 42 87.3 52 68.8 74 54.4 91 82.1 66 28.4 71 45.0 116 42.5 95 36.8 96 56.3 121 50.0 40
Aniso. Huber-L1 [22]84.5 36.7 123 33.5 104 48.6 126 28.5 134 34.3 129 38.2 109 29.3 84 31.8 86 38.9 52 52.5 71 47.5 77 80.6 75 72.0 43 66.7 42 87.4 78 68.6 36 54.3 86 81.9 28 28.5 97 45.0 116 42.4 32 36.8 96 56.0 98 50.3 107
RFlow [88]86.0 36.2 59 33.0 85 48.2 93 27.6 109 33.7 114 37.1 31 29.3 84 32.5 103 39.2 97 52.6 87 47.8 94 80.6 75 72.0 43 66.8 54 87.3 52 68.6 36 53.8 51 81.9 28 28.5 97 45.5 138 42.6 135 37.2 143 56.9 149 50.3 107
EpicFlow [100]86.0 36.2 59 33.3 95 48.1 75 26.9 80 33.1 97 37.5 75 29.4 99 33.0 118 39.0 71 52.6 87 48.5 132 80.8 116 72.3 111 67.0 82 87.5 140 68.6 36 54.1 73 82.0 38 28.4 71 44.8 103 42.4 32 36.6 72 55.7 81 50.4 121
CtxSyn [134]86.9 26.8 4 25.3 9 36.2 4 23.8 7 27.2 7 39.5 126 26.3 20 26.1 22 48.0 185 51.4 20 44.1 21 82.1 172 71.6 34 66.0 34 87.1 35 70.2 180 54.2 81 84.5 194 39.3 196 45.6 142 77.2 195 38.0 174 52.3 31 59.7 193
DMF_ROB [135]87.2 36.2 59 32.6 70 48.1 75 27.4 102 33.5 109 37.7 90 30.2 150 34.4 149 39.6 131 52.7 98 47.6 81 80.6 75 72.1 58 66.7 42 87.6 175 68.3 30 53.2 36 82.0 38 28.6 116 44.1 68 43.0 164 36.3 41 55.1 48 50.2 87
Brox et al. [5]87.7 36.3 78 32.4 60 48.2 93 27.8 117 34.1 125 38.0 106 29.8 130 33.9 140 39.6 131 52.5 71 47.0 62 80.4 48 72.2 76 66.9 67 87.5 140 68.7 52 53.8 51 82.1 66 28.4 71 44.9 108 42.5 95 36.5 60 55.5 70 50.2 87
SepConv-v1 [125]88.1 27.1 5 27.5 19 36.4 5 24.9 11 31.0 45 40.3 135 27.6 25 28.4 27 48.9 189 54.0 164 47.3 68 83.0 186 72.0 43 66.7 42 87.1 35 69.1 118 52.6 32 83.6 183 32.2 190 43.9 56 55.7 191 37.0 122 53.8 34 55.4 190
FESL [72]88.3 36.6 112 33.9 127 48.0 59 26.4 56 31.7 62 37.3 53 29.1 65 31.3 71 38.9 52 52.6 87 47.6 81 80.3 33 72.4 145 67.3 150 87.4 78 69.3 143 55.9 164 82.1 66 28.4 71 44.9 108 42.3 15 37.0 122 56.6 137 50.1 63
PWC-Net_RVC [143]88.5 36.4 91 34.4 146 48.0 59 27.3 99 33.9 121 37.7 90 29.7 122 34.8 158 39.1 81 52.5 71 48.9 150 80.4 48 72.4 145 67.3 150 87.4 78 69.0 104 54.1 73 82.2 97 28.2 36 43.9 56 42.4 32 36.3 41 55.1 48 49.9 28
S2D-Matching [83]88.6 36.6 112 34.2 141 48.2 93 26.9 80 32.5 82 37.2 46 28.8 43 30.7 53 38.9 52 52.1 43 46.4 44 80.9 127 72.3 111 67.1 114 87.5 140 69.1 118 55.3 142 82.2 97 28.5 97 44.7 94 42.4 32 36.7 87 55.9 92 50.2 87
p-harmonic [29]89.7 35.9 36 32.1 53 47.9 42 28.2 125 34.3 129 37.8 96 29.4 99 34.2 145 39.4 113 53.0 124 47.7 87 80.7 100 72.2 76 67.0 82 87.3 52 68.8 74 54.1 73 82.3 122 28.5 97 45.5 138 42.4 32 36.6 72 56.0 98 50.2 87
LiteFlowNet [138]89.8 36.4 91 34.2 141 48.0 59 27.1 92 33.4 107 37.5 75 29.6 113 35.1 166 39.1 81 53.3 139 50.7 177 80.6 75 72.1 58 66.9 67 87.3 52 69.1 118 55.1 132 82.0 38 28.6 116 45.5 138 42.3 15 36.1 34 54.8 43 49.9 28
ComplOF-FED-GPU [35]91.0 36.3 78 33.4 100 48.0 59 26.8 73 33.0 94 37.3 53 30.4 155 34.0 141 39.6 131 52.5 71 48.1 111 80.9 127 72.1 58 66.8 54 87.4 78 68.7 52 54.3 86 82.1 66 28.5 97 45.1 124 42.5 95 36.8 96 56.0 98 50.2 87
TC-Flow [46]91.7 36.2 59 33.2 92 48.2 93 26.9 80 33.5 109 37.5 75 29.5 109 33.6 134 38.9 52 52.1 43 47.1 65 80.6 75 72.3 111 67.2 134 87.5 140 69.0 104 54.8 118 82.3 122 28.4 71 44.4 78 42.5 95 36.6 72 56.1 107 50.1 63
ProFlow_ROB [142]91.8 36.2 59 32.8 78 48.2 93 26.9 80 33.3 103 37.7 90 29.1 65 32.2 94 38.8 35 52.6 87 48.7 143 80.7 100 72.4 145 67.2 134 87.4 78 68.7 52 53.8 51 82.2 97 28.5 97 45.2 129 42.4 32 37.0 122 56.5 131 50.3 107
DPOF [18]92.5 36.7 123 34.5 148 48.6 126 26.1 33 30.6 34 37.6 85 29.8 130 31.4 72 39.3 107 52.8 110 48.6 136 80.8 116 72.0 43 66.8 54 87.3 52 69.1 118 55.3 142 81.9 28 28.5 97 44.5 86 42.5 95 36.9 113 56.5 131 50.0 40
EPPM w/o HM [86]92.7 35.8 30 32.3 58 47.6 20 26.7 67 33.0 94 36.9 10 30.0 141 35.5 170 39.4 113 52.6 87 48.9 150 80.4 48 72.2 76 67.1 114 87.4 78 69.3 143 55.9 164 82.3 122 28.4 71 44.9 108 42.5 95 36.8 96 56.1 107 50.1 63
JOF [136]93.1 36.9 136 34.2 141 48.7 131 26.0 28 30.3 26 37.3 53 28.8 43 30.2 36 38.9 52 52.3 56 46.7 53 81.0 139 72.4 145 67.2 134 87.5 140 69.2 131 55.5 152 82.2 97 28.2 36 43.8 55 42.5 95 36.9 113 56.3 121 50.4 121
OFH [38]94.4 36.4 91 33.8 122 48.2 93 27.4 102 33.3 103 37.4 65 29.7 122 35.0 162 39.0 71 52.5 71 48.3 124 80.9 127 72.2 76 67.0 82 87.4 78 68.7 52 54.2 81 82.1 66 28.6 116 45.4 133 42.5 95 36.6 72 56.0 98 50.1 63
Efficient-NL [60]94.7 36.5 103 33.6 112 48.0 59 26.7 67 32.0 67 37.1 31 29.9 137 31.4 72 39.3 107 52.7 98 47.7 87 80.4 48 72.2 76 67.0 82 87.3 52 69.5 154 57.0 184 81.9 28 28.6 116 45.9 148 42.4 32 37.9 170 58.1 179 50.1 63
PMF [73]97.3 35.9 36 32.0 50 47.7 27 26.9 80 33.5 109 36.9 10 29.6 113 34.5 151 39.1 81 52.5 71 47.8 94 80.4 48 72.5 161 67.5 169 87.4 78 69.2 131 55.0 130 82.4 136 28.5 97 45.0 116 42.5 95 37.3 150 57.3 162 50.0 40
Local-TV-L1 [65]98.2 37.5 153 33.0 85 49.7 159 29.3 153 34.5 137 40.3 135 29.2 73 31.6 79 39.1 81 53.3 139 47.3 68 83.1 189 72.1 58 66.9 67 87.4 78 69.3 143 53.4 40 83.2 174 28.2 36 43.9 56 42.4 32 36.4 52 55.1 48 50.4 121
OAR-Flow [123]98.5 36.5 103 33.0 85 48.4 114 27.0 87 33.0 94 37.8 96 29.2 73 33.3 127 38.9 52 52.1 43 47.6 81 80.5 67 72.4 145 67.3 150 87.6 175 68.9 88 54.5 98 82.2 97 28.5 97 44.7 94 42.4 32 36.9 113 56.5 131 50.4 121
LFNet_ROB [145]98.5 36.6 112 33.5 104 48.4 114 28.4 131 35.0 150 38.4 113 29.9 137 34.8 158 39.5 120 52.4 63 47.8 94 80.6 75 72.1 58 66.9 67 87.4 78 68.9 88 54.8 118 82.1 66 28.3 48 44.4 78 42.6 135 36.6 72 55.4 64 50.4 121
Sparse Occlusion [54]98.5 36.5 103 33.7 116 48.2 93 27.6 109 34.1 125 37.3 53 29.3 84 31.8 86 38.8 35 52.8 110 48.1 111 80.5 67 72.3 111 67.1 114 87.4 78 69.2 131 56.1 171 82.0 38 28.5 97 45.4 133 42.3 15 37.2 143 57.0 155 50.2 87
OFRI [154]98.6 38.3 170 27.4 18 53.7 186 31.4 173 31.8 66 56.5 198 27.8 26 25.4 20 54.3 198 51.1 16 43.1 14 81.2 155 64.8 18 58.1 18 83.8 19 68.4 31 51.8 31 83.4 180 41.6 198 42.9 33 81.5 199 38.1 176 50.1 25 63.8 198
TC/T-Flow [77]98.9 36.6 112 33.7 116 47.9 42 26.8 73 32.9 91 37.1 31 29.1 65 31.9 91 38.8 35 52.7 98 48.5 132 80.4 48 72.5 161 67.4 162 87.5 140 69.1 118 55.2 137 82.1 66 28.6 116 45.4 133 42.5 95 37.0 122 56.9 149 50.0 40
SRR-TVOF-NL [89]99.4 36.6 112 33.5 104 48.2 93 27.7 113 34.3 129 37.9 102 29.5 109 33.2 123 39.1 81 53.1 129 48.1 111 80.2 26 72.2 76 67.1 114 87.3 52 68.9 88 55.7 160 81.8 26 28.5 97 44.9 108 42.4 32 37.5 160 58.0 177 50.1 63
TF+OM [98]99.5 36.3 78 33.0 85 48.5 118 26.9 80 32.2 73 39.2 122 28.6 29 32.4 99 38.9 52 52.8 110 48.2 122 80.7 100 72.3 111 67.1 114 87.4 78 69.0 104 54.5 98 82.3 122 28.4 71 45.1 124 42.5 95 37.0 122 56.4 128 50.6 152
ALD-Flow [66]101.8 36.7 123 33.9 127 48.6 126 27.0 87 33.2 99 37.9 102 29.3 84 33.4 130 38.9 52 52.5 71 48.0 104 80.9 127 72.4 145 67.2 134 87.6 175 68.9 88 54.4 91 82.2 97 28.2 36 43.6 46 42.4 32 37.0 122 56.6 137 50.3 107
CLG-TV [48]101.8 36.6 112 33.4 100 48.5 118 28.2 125 34.4 135 38.2 109 29.7 122 33.6 134 39.4 113 52.8 110 48.0 104 80.9 127 72.2 76 66.9 67 87.5 140 68.7 52 54.0 66 82.1 66 28.4 71 45.1 124 42.4 32 37.0 122 56.5 131 50.2 87
MS-PFT [159]102.1 34.5 19 30.4 33 45.4 14 32.6 183 34.0 123 49.8 190 30.1 147 30.6 48 51.8 196 56.4 188 51.2 182 82.1 172 64.8 18 58.1 18 84.7 24 66.8 25 44.8 20 84.2 191 40.2 197 43.3 38 80.1 198 34.3 22 45.9 12 59.7 193
SIOF [67]102.3 36.7 123 34.1 136 48.2 93 29.1 148 35.4 159 39.7 129 29.4 99 32.9 112 39.1 81 52.7 98 47.7 87 80.9 127 71.9 38 66.6 36 87.4 78 69.1 118 54.3 86 82.4 136 28.3 48 44.6 89 42.4 32 37.3 150 56.8 145 50.3 107
SimpleFlow [49]102.8 36.5 103 34.2 141 48.2 93 27.2 96 32.8 88 37.3 53 30.1 147 31.7 81 39.4 113 52.0 35 46.3 41 80.7 100 72.3 111 67.2 134 87.4 78 69.0 104 55.4 148 82.0 38 28.7 134 47.1 164 42.6 135 37.0 122 56.8 145 50.1 63
AggregFlow [95]102.9 37.1 142 34.8 156 48.5 118 27.3 99 33.2 99 38.1 107 28.7 37 30.2 36 38.5 17 52.9 120 48.6 136 80.3 33 72.4 145 67.2 134 87.6 175 69.3 143 54.5 98 82.6 155 28.3 48 44.2 71 42.5 95 36.7 87 56.0 98 50.4 121
ContinualFlow_ROB [148]103.0 37.6 156 36.8 178 49.1 141 28.6 137 35.6 161 40.5 138 30.4 155 36.3 175 39.5 120 52.8 110 49.3 159 80.6 75 72.3 111 67.3 150 87.4 78 68.4 31 54.0 66 81.9 28 28.3 48 44.4 78 42.3 15 36.3 41 55.9 92 49.9 28
Complementary OF [21]103.2 36.1 51 33.3 95 47.8 31 26.7 67 33.2 99 37.3 53 30.4 155 32.9 112 39.5 120 52.8 110 48.7 143 81.1 150 72.3 111 67.2 134 87.3 52 68.8 74 54.7 110 82.2 97 28.7 134 45.6 142 42.5 95 36.8 96 56.7 140 50.3 107
SuperSlomo [130]103.5 32.1 13 27.9 20 42.8 10 29.9 165 32.2 73 48.3 189 31.1 173 30.8 57 49.0 190 53.4 144 45.1 26 82.8 183 70.9 31 65.3 31 86.6 31 69.5 154 51.7 30 84.0 190 32.2 190 42.2 27 55.7 191 37.2 143 52.2 30 57.3 192
F-TV-L1 [15]104.7 37.4 151 34.6 150 49.2 144 28.8 143 34.9 148 38.3 111 29.7 122 34.1 143 39.5 120 52.7 98 47.6 81 81.0 139 71.7 35 66.5 35 87.4 78 68.8 74 53.5 44 82.4 136 28.3 48 44.3 74 42.4 32 37.1 134 56.3 121 50.6 152
LDOF [28]105.0 37.1 142 33.7 116 48.8 135 29.5 156 35.3 157 40.6 142 30.0 141 34.3 147 39.7 136 52.8 110 47.9 99 80.9 127 72.2 76 66.9 67 87.4 78 68.8 74 53.6 46 82.3 122 28.3 48 44.5 86 42.4 32 36.6 72 55.8 89 50.4 121
MLDP_OF [87]105.4 36.2 59 32.9 83 48.0 59 27.0 87 32.7 86 37.2 46 29.1 65 31.8 86 38.8 35 52.6 87 47.3 68 80.8 116 72.3 111 67.1 114 87.5 140 70.5 188 56.6 180 83.6 183 28.6 116 44.8 103 42.8 155 36.9 113 56.1 107 50.5 142
MCPFlow_RVC [197]106.3 37.1 142 35.2 164 48.5 118 27.5 106 33.5 109 39.1 120 29.0 55 32.8 111 38.6 26 52.8 110 48.5 132 80.6 75 72.6 174 67.6 173 87.4 78 69.1 118 56.3 178 82.1 66 28.2 36 43.6 46 42.5 95 36.9 113 59.8 191 49.5 15
Classic++ [32]107.0 36.4 91 33.5 104 48.4 114 27.4 102 33.7 114 37.6 85 29.6 113 33.6 134 39.2 97 52.7 98 47.3 68 80.9 127 72.2 76 67.0 82 87.5 140 69.1 118 54.5 98 82.5 147 28.5 97 44.9 108 42.6 135 36.8 96 56.2 116 50.3 107
LSM_FLOW_RVC [182]107.0 37.2 148 36.5 175 48.6 126 28.9 145 36.7 173 39.3 124 29.8 130 36.2 174 39.0 71 53.0 124 50.3 172 80.4 48 72.2 76 67.0 82 87.3 52 68.7 52 54.0 66 82.1 66 28.7 134 45.9 148 42.4 32 36.6 72 55.3 57 50.4 121
IAOF [50]107.5 38.0 168 34.2 141 49.8 160 31.7 176 37.9 178 41.1 147 28.9 51 32.6 105 39.4 113 53.7 154 48.1 111 80.8 116 72.0 43 66.7 42 87.5 140 68.9 88 54.1 73 82.2 97 28.3 48 45.1 124 42.3 15 36.8 96 56.1 107 50.2 87
C-RAFT_RVC [181]108.2 37.9 162 36.6 176 49.1 141 28.0 121 34.7 140 39.2 122 29.7 122 33.5 133 39.2 97 53.1 129 49.6 164 80.6 75 72.2 76 67.0 82 87.2 43 69.0 104 55.1 132 82.0 38 28.4 71 44.5 86 42.5 95 36.7 87 55.7 81 50.4 121
CostFilter [40]108.7 35.9 36 32.7 75 47.6 20 26.8 73 33.5 109 37.1 31 29.7 122 35.6 172 39.2 97 52.9 120 49.4 161 80.3 33 72.6 174 67.6 173 87.4 78 69.6 160 54.8 118 83.1 173 28.6 116 45.6 142 42.6 135 37.0 122 56.7 140 49.9 28
Fusion [6]109.0 36.0 43 32.7 75 47.8 31 26.8 73 32.1 71 37.5 75 29.5 109 31.5 77 39.5 120 53.5 145 48.6 136 80.7 100 72.6 174 68.0 186 87.1 35 69.3 143 57.6 191 81.8 26 28.7 134 47.1 164 42.5 95 38.2 179 59.9 193 50.0 40
FlowNetS+ft+v [110]109.4 36.8 130 33.0 85 48.7 131 29.5 156 35.6 161 40.5 138 29.8 130 34.3 147 39.5 120 52.8 110 48.2 122 80.8 116 72.2 76 67.0 82 87.4 78 68.7 52 53.9 61 82.1 66 28.6 116 45.9 148 42.5 95 36.7 87 56.0 98 50.4 121
Shiralkar [42]109.6 36.5 103 34.6 150 48.1 75 28.3 129 34.3 129 37.2 46 29.8 130 36.9 181 40.0 147 53.9 160 49.0 153 80.5 67 71.8 36 66.6 36 87.2 43 69.2 131 55.1 132 82.4 136 29.2 161 48.0 175 42.5 95 36.6 72 55.7 81 50.1 63
SVFilterOh [109]109.8 36.3 78 32.2 55 48.1 75 26.2 45 30.9 42 37.4 65 29.2 73 30.6 48 39.3 107 52.6 87 47.5 77 81.0 139 72.6 174 67.6 173 87.6 175 69.3 143 55.9 164 82.3 122 28.5 97 43.7 52 43.3 172 37.3 150 57.1 158 51.0 163
Occlusion-TV-L1 [63]110.1 36.6 112 33.8 122 48.5 118 28.4 131 34.8 145 37.7 90 29.5 109 33.0 118 39.5 120 53.0 124 48.1 111 81.1 150 72.1 58 66.8 54 87.5 140 68.9 88 53.4 40 82.4 136 29.0 154 44.7 94 42.6 135 36.8 96 55.6 76 50.4 121
TriFlow [93]110.2 37.0 139 35.3 165 48.8 135 28.7 141 34.5 137 41.0 145 29.2 73 33.4 130 38.8 35 53.0 124 48.8 149 80.4 48 72.3 111 67.3 150 87.4 78 69.2 131 55.5 152 82.1 66 28.5 97 44.8 103 42.4 32 36.9 113 56.4 128 50.1 63
TOF-M [150]110.7 31.5 10 27.3 17 42.1 9 28.6 137 32.8 88 46.1 184 31.1 173 31.9 91 50.6 195 53.6 149 46.3 41 82.8 183 71.2 33 65.6 33 87.1 35 70.3 185 53.7 48 85.0 195 38.0 194 43.3 38 76.6 194 39.1 190 54.5 38 61.2 197
3DFlow [133]111.5 36.4 91 33.8 122 47.9 42 26.5 60 32.1 71 37.1 31 29.8 130 31.7 81 39.1 81 52.5 71 47.7 87 80.6 75 72.5 161 67.3 150 88.0 189 70.0 169 57.8 192 82.2 97 29.1 159 47.4 168 42.5 95 37.4 157 57.6 169 49.9 28
CRTflow [81]111.6 36.7 123 33.8 122 48.5 118 27.7 113 33.8 117 37.4 65 30.7 165 35.3 167 40.9 165 52.9 120 48.1 111 81.8 169 72.2 76 66.9 67 87.4 78 68.9 88 54.1 73 82.3 122 28.4 71 44.9 108 42.5 95 36.8 96 56.1 107 50.5 142
CNN-flow-warp+ref [115]111.7 36.3 78 31.7 42 48.7 131 28.5 134 34.7 140 39.5 126 30.4 155 35.0 162 39.8 140 54.0 164 48.1 111 81.2 155 72.3 111 67.0 82 87.4 78 68.6 36 53.2 36 82.4 136 28.8 144 47.1 164 42.5 95 36.6 72 55.7 81 50.3 107
FLAVR [188]113.2 45.2 195 36.2 174 56.9 193 40.6 197 38.1 180 54.5 195 32.0 185 32.9 112 47.2 184 63.8 197 58.4 196 80.8 116 63.8 6 57.1 9 82.1 1 63.7 6 41.8 8 79.8 5 32.0 189 44.3 74 48.0 189 31.9 10 44.0 8 50.2 87
CVENG22+RIC [199]113.6 36.6 112 34.1 136 48.3 113 27.4 102 34.0 123 37.8 96 29.6 113 34.2 145 39.2 97 53.1 129 49.5 163 80.9 127 72.3 111 67.1 114 87.4 78 68.9 88 54.7 110 82.2 97 28.6 116 45.6 142 42.4 32 37.0 122 56.7 140 50.4 121
FlowNet2 [120]114.8 39.4 179 38.2 185 50.4 168 29.2 151 34.8 145 41.9 162 30.0 141 34.6 153 39.4 113 53.3 139 51.0 181 80.6 75 72.5 161 67.4 162 87.4 78 68.8 74 54.3 86 82.0 38 28.4 71 45.0 116 42.3 15 36.5 60 55.7 81 49.8 21
TCOF [69]114.8 36.6 112 33.9 127 48.1 75 29.1 148 35.7 164 38.3 111 29.0 55 31.4 72 38.7 28 52.8 110 48.7 143 80.6 75 72.2 76 67.1 114 87.4 78 69.3 143 56.0 167 82.1 66 28.7 134 46.2 154 42.5 95 38.2 179 58.7 188 50.5 142
Adaptive [20]115.2 36.8 130 34.4 146 48.5 118 28.8 143 35.2 156 37.7 90 29.4 99 33.2 123 39.2 97 52.6 87 47.6 81 80.6 75 72.3 111 67.0 82 87.5 140 69.1 118 54.7 110 82.3 122 28.7 134 46.0 151 42.4 32 37.3 150 56.9 149 50.4 121
ResPWCR_ROB [140]115.2 36.4 91 34.0 132 48.1 75 28.2 125 34.9 148 38.9 118 30.7 165 35.0 162 39.7 136 53.7 154 50.6 174 81.5 163 71.8 36 66.6 36 87.0 34 70.7 191 56.0 167 84.3 192 28.4 71 44.9 108 42.5 95 36.5 60 55.9 92 50.0 40
AugFNG_ROB [139]115.7 37.7 158 35.6 168 49.6 154 29.5 156 36.0 166 41.4 153 30.4 155 37.7 185 40.1 152 53.5 145 50.5 173 81.0 139 72.5 161 67.5 169 87.4 78 68.7 52 54.4 91 82.0 38 28.5 97 44.4 78 42.4 32 35.3 27 54.0 36 49.4 14
Modified CLG [34]115.8 36.9 136 32.8 78 49.4 148 30.9 172 36.3 171 42.8 164 30.0 141 34.8 158 39.9 143 53.0 124 47.9 99 80.7 100 72.2 76 66.9 67 87.5 140 68.7 52 53.8 51 82.2 97 28.4 71 45.1 124 42.5 95 36.9 113 56.2 116 50.5 142
EPMNet [131]116.5 38.9 175 38.5 188 49.9 163 29.0 147 34.2 128 41.2 148 30.0 141 34.6 153 39.4 113 53.9 160 52.7 190 80.6 75 72.5 161 67.4 162 87.4 78 69.0 104 55.5 152 82.0 38 28.4 71 45.0 116 42.3 15 36.3 41 55.3 57 49.8 21
IIOF-NLDP [129]117.0 36.3 78 33.3 95 47.7 27 27.6 109 34.3 129 37.4 65 29.8 130 31.7 81 39.2 97 53.3 139 48.7 143 81.2 155 72.2 76 67.0 82 87.5 140 69.8 164 56.8 181 82.2 97 29.4 166 50.8 193 42.8 155 37.1 134 56.8 145 49.9 28
Steered-L1 [116]118.6 36.0 43 32.9 83 47.9 42 27.0 87 33.3 103 37.7 90 30.3 153 32.3 96 39.9 143 53.2 133 48.0 104 81.0 139 72.5 161 67.5 169 87.5 140 68.9 88 55.0 130 82.2 97 28.8 144 46.7 161 42.7 149 37.0 122 57.3 162 50.3 107
Nguyen [33]120.5 39.6 180 33.9 127 52.6 184 32.5 181 37.9 178 43.3 169 30.0 141 35.5 170 40.2 153 54.1 169 49.0 153 80.9 127 72.0 43 66.8 54 87.4 78 68.6 36 53.8 51 82.0 38 28.8 144 47.8 173 42.4 32 36.8 96 56.1 107 50.3 107
StereoOF-V1MT [117]120.7 36.8 130 35.3 165 48.1 75 28.3 129 35.1 153 36.9 10 31.4 177 36.6 177 40.5 156 54.6 179 48.6 136 81.3 159 72.0 43 66.8 54 87.2 43 69.5 154 54.9 122 82.6 155 29.7 178 48.8 182 42.7 149 36.5 60 55.1 48 50.1 63
BriefMatch [122]120.8 36.3 78 33.3 95 48.0 59 27.2 96 33.4 107 38.5 117 30.6 163 32.6 105 40.6 158 54.0 164 48.6 136 82.8 183 72.4 145 67.3 150 87.3 52 70.2 180 55.6 158 83.9 187 28.3 48 44.3 74 42.7 149 36.6 72 55.7 81 50.5 142
CompactFlow_ROB [155]121.0 37.2 148 35.3 165 48.8 135 28.9 145 35.6 161 41.4 153 30.3 153 36.8 179 39.1 81 53.5 145 50.7 177 81.0 139 72.2 76 67.0 82 87.4 78 69.2 131 56.0 167 82.0 38 28.6 116 45.8 147 42.4 32 36.8 96 56.0 98 50.1 63
SPSA-learn [13]124.8 37.4 151 33.6 112 49.4 148 29.8 163 35.1 153 41.4 153 30.9 169 33.2 123 40.7 159 53.5 145 47.2 66 80.4 48 72.2 76 67.0 82 87.4 78 68.8 74 54.1 73 82.2 97 29.5 171 52.2 196 42.9 161 37.1 134 57.0 155 50.3 107
GraphCuts [14]125.2 38.0 168 35.1 161 49.5 153 28.4 131 33.9 121 41.3 150 31.3 176 30.8 57 40.7 159 53.7 154 48.3 124 81.0 139 72.1 58 67.1 114 87.1 35 68.6 36 54.9 122 81.7 23 28.8 144 46.3 155 42.8 155 37.7 164 58.5 184 50.4 121
Dynamic MRF [7]126.2 36.2 59 34.1 136 48.0 59 27.5 106 34.6 139 37.4 65 30.9 169 36.8 179 40.4 155 54.5 177 49.3 159 81.9 170 71.9 38 66.8 54 87.2 43 69.4 152 55.5 152 82.5 147 29.0 154 47.8 173 42.5 95 37.5 160 56.8 145 50.5 142
ROF-ND [105]126.7 37.0 139 32.8 78 48.1 75 27.7 113 34.7 140 37.6 85 29.7 122 32.3 96 39.1 81 54.2 172 51.4 183 80.4 48 72.4 145 67.3 150 87.4 78 69.5 154 56.9 182 82.0 38 29.7 178 49.0 185 43.2 170 37.8 168 57.7 173 50.2 87
IRR-PWC_RVC [180]127.0 38.3 170 37.5 180 49.6 154 29.2 151 35.5 160 41.6 158 30.5 160 38.0 188 39.5 120 54.1 169 52.1 186 80.7 100 72.6 174 67.5 169 87.4 78 69.1 118 55.4 148 82.1 66 28.4 71 44.8 103 42.3 15 36.6 72 56.3 121 49.6 17
HBpMotionGpu [43]127.3 38.8 174 35.9 170 50.9 177 32.1 178 38.2 181 44.4 175 29.2 73 31.7 81 39.3 107 53.9 160 49.6 164 81.5 163 72.1 58 67.0 82 87.1 35 69.5 154 54.9 122 82.4 136 28.3 48 44.4 78 42.5 95 37.3 150 56.5 131 51.1 164
2D-CLG [1]127.8 37.9 162 33.5 104 50.5 171 32.5 181 37.4 175 45.0 178 30.8 168 34.8 158 40.7 159 53.7 154 48.3 124 80.5 67 72.3 111 67.1 114 87.6 175 68.6 36 53.2 36 82.2 97 28.8 144 46.7 161 42.5 95 36.9 113 55.6 76 50.3 107
Black & Anandan [4]127.8 37.9 162 34.1 136 49.6 154 30.7 170 36.0 166 41.2 148 31.0 171 34.7 156 40.3 154 53.9 160 48.6 136 80.7 100 72.3 111 67.0 82 87.4 78 69.0 104 53.8 51 82.5 147 28.8 144 46.5 156 42.4 32 37.0 122 56.1 107 50.4 121
TV-L1-improved [17]128.0 36.6 112 34.1 136 48.4 114 28.6 137 35.1 153 37.8 96 30.5 160 33.2 123 40.0 147 52.7 98 48.0 104 80.9 127 72.3 111 67.2 134 87.4 78 69.1 118 54.9 122 82.3 122 28.8 144 47.3 167 42.6 135 37.2 143 56.7 140 50.6 152
CBF [12]129.5 36.4 91 32.5 65 48.9 138 27.5 106 33.8 117 37.9 102 29.3 84 31.6 79 39.1 81 53.2 133 48.1 111 82.6 179 72.4 145 67.2 134 87.7 185 69.2 131 55.3 142 82.3 122 28.7 134 46.1 152 42.9 161 37.9 170 57.7 173 51.7 173
TVL1_RVC [175]130.6 39.9 183 35.0 158 52.5 183 33.3 185 38.7 185 45.0 178 29.7 122 34.1 143 39.8 140 54.0 164 48.1 111 81.2 155 72.2 76 67.0 82 87.4 78 69.0 104 53.6 46 82.5 147 28.7 134 46.5 156 42.5 95 36.8 96 55.9 92 50.4 121
Correlation Flow [76]131.9 36.2 59 33.4 100 47.7 27 27.7 113 34.3 129 37.3 53 29.4 99 31.4 72 38.8 35 53.1 129 48.5 132 81.3 159 72.8 183 67.6 173 88.6 197 70.1 172 57.1 185 82.6 155 29.4 166 48.8 182 43.0 164 37.7 164 57.9 176 50.5 142
Rannacher [23]132.3 36.7 123 34.5 148 48.7 131 28.7 141 35.3 157 38.1 107 30.5 160 34.0 141 39.9 143 52.7 98 48.0 104 80.8 116 72.4 145 67.2 134 87.5 140 69.0 104 54.6 105 82.3 122 28.8 144 47.0 163 42.6 135 37.1 134 56.4 128 50.6 152
UnFlow [127]133.0 39.2 178 37.9 182 50.6 174 32.3 179 38.9 188 41.3 150 31.6 182 38.5 189 40.8 164 53.2 133 48.7 143 80.9 127 72.0 43 66.7 42 87.4 78 69.5 154 54.6 105 82.4 136 28.2 36 43.2 37 42.4 32 39.3 193 58.3 181 51.2 165
HBM-GC [103]133.4 37.7 158 34.7 153 49.8 160 27.1 92 32.4 80 37.9 102 28.8 43 29.6 30 39.2 97 52.6 87 47.3 68 80.8 116 73.2 189 68.1 187 88.2 192 70.0 169 57.3 187 82.7 159 28.9 153 45.0 116 43.5 175 37.6 162 57.2 160 51.3 167
TriangleFlow [30]133.8 37.0 139 34.9 157 48.5 118 28.0 121 34.7 140 37.5 75 30.2 150 33.0 118 39.9 143 53.2 133 49.0 153 81.1 150 72.0 43 66.9 67 87.1 35 69.8 164 56.1 171 82.4 136 29.2 161 48.5 179 42.8 155 38.1 176 58.5 184 50.5 142
SegOF [10]135.7 37.6 156 33.2 92 50.0 164 29.1 148 34.7 140 41.0 145 31.4 177 35.3 167 40.7 159 53.6 149 50.7 177 80.6 75 72.3 111 67.2 134 87.5 140 69.0 104 55.3 142 82.2 97 29.0 154 48.6 180 42.7 149 36.7 87 55.8 89 50.4 121
WRT [146]136.4 36.6 112 34.0 132 47.9 42 28.1 123 33.8 117 37.5 75 31.1 173 31.4 72 39.6 131 53.2 133 48.4 130 80.8 116 72.7 181 67.7 181 87.8 187 70.1 172 58.7 195 82.2 97 29.8 181 53.4 198 42.9 161 37.6 162 58.4 182 49.8 21
BlockOverlap [61]136.8 38.5 173 33.2 92 51.3 178 30.0 167 34.4 135 42.8 164 29.4 99 30.4 44 40.0 147 53.2 133 46.9 58 83.0 186 72.9 186 67.6 173 88.3 193 69.7 162 54.1 73 83.3 179 28.7 134 44.1 68 43.5 175 37.1 134 55.3 57 51.8 174
IAOF2 [51]138.5 37.9 162 35.9 170 49.1 141 29.6 160 36.1 169 40.0 132 29.3 84 33.4 130 40.0 147 54.1 169 50.2 171 81.0 139 72.4 145 67.4 162 87.4 78 69.2 131 54.9 122 82.4 136 28.6 116 45.5 138 42.4 32 37.9 170 57.6 169 50.6 152
Ad-TV-NDC [36]140.0 40.4 186 35.1 161 53.1 185 31.9 177 36.7 173 43.8 173 29.4 99 32.9 112 39.1 81 54.5 177 49.2 158 82.1 172 72.5 161 67.3 150 87.5 140 69.3 143 53.9 61 82.7 159 28.6 116 45.4 133 42.4 32 37.2 143 56.2 116 50.6 152
OFRF [132]145.2 38.9 175 36.1 173 50.4 168 29.5 156 35.0 150 40.5 138 29.6 113 34.4 149 39.0 71 53.3 139 48.9 150 81.1 150 72.6 174 67.7 181 87.3 52 70.1 172 57.3 187 82.5 147 29.1 159 47.6 171 42.6 135 37.4 157 58.0 177 50.0 40
LocallyOriented [52]148.8 37.5 153 35.9 170 49.2 144 29.6 160 36.2 170 39.1 120 30.1 147 33.8 138 39.5 120 53.7 154 50.0 169 81.3 159 72.3 111 67.2 134 87.5 140 70.2 180 56.2 177 82.9 168 28.8 144 45.6 142 42.5 95 37.7 164 57.6 169 50.5 142
ACK-Prior [27]149.2 36.4 91 33.7 116 48.1 75 26.7 67 33.1 97 37.1 31 30.7 165 33.3 127 39.7 136 53.6 149 50.0 169 81.0 139 73.5 193 68.6 190 88.3 193 70.8 192 59.8 196 82.7 159 29.7 178 48.7 181 43.6 179 39.5 194 62.1 196 51.3 167
AdaConv-v1 [124]149.6 37.2 148 36.6 176 47.5 18 34.3 189 39.1 190 51.1 193 36.1 194 39.4 191 52.9 197 58.2 192 53.1 191 83.8 191 70.9 31 65.4 32 86.6 31 69.7 162 54.7 110 84.4 193 38.6 195 46.5 156 77.4 196 38.2 179 54.6 40 60.3 196
Horn & Schunck [3]151.0 37.9 162 35.1 161 49.6 154 31.4 173 37.7 177 41.8 160 31.7 183 37.4 184 41.5 167 55.8 183 50.6 174 81.3 159 72.2 76 67.0 82 87.4 78 69.2 131 54.2 81 82.7 159 29.5 171 48.9 184 42.6 135 37.8 168 57.2 160 50.9 162
StereoFlow [44]151.3 46.3 197 45.9 198 54.3 188 38.3 196 45.4 198 45.7 182 29.3 84 33.8 138 39.1 81 52.9 120 47.7 87 81.0 139 74.4 197 70.5 198 87.6 175 72.0 196 66.3 198 82.4 136 28.4 71 45.0 116 42.4 32 38.0 174 59.1 189 50.5 142
WOLF_ROB [144]151.3 38.3 170 38.0 183 49.0 140 29.6 160 35.9 165 39.0 119 30.6 163 35.0 162 39.8 140 54.4 176 52.4 189 82.0 171 72.5 161 67.6 173 87.4 78 69.8 164 55.4 148 82.8 165 29.4 166 49.1 186 42.5 95 37.1 134 56.6 137 50.2 87
Filter Flow [19]152.8 37.8 161 34.6 150 49.8 160 30.8 171 36.0 166 44.3 174 29.4 99 32.4 99 39.5 120 54.2 172 48.1 111 82.2 175 72.7 181 67.7 181 87.6 175 69.2 131 55.1 132 82.5 147 28.7 134 46.5 156 42.6 135 38.3 185 58.4 182 51.4 170
TI-DOFE [24]154.6 42.0 189 37.5 180 54.8 191 35.2 191 41.1 195 46.8 186 31.4 177 37.7 185 41.6 169 56.1 185 50.6 174 81.6 165 72.0 43 66.9 67 87.2 43 69.4 152 54.4 91 82.6 155 29.2 161 47.6 171 42.6 135 38.2 179 57.5 167 50.8 160
SILK [80]156.9 39.6 180 38.1 184 51.5 181 32.4 180 38.5 184 43.6 172 32.4 186 37.2 183 41.5 167 55.4 181 49.7 166 83.0 186 72.2 76 67.0 82 87.4 78 70.0 169 54.7 110 83.4 180 29.0 154 46.5 156 42.8 155 37.4 157 56.7 140 50.7 158
Bartels [41]157.7 37.1 142 35.0 158 49.3 147 28.2 125 34.8 145 40.5 138 29.9 137 33.1 121 40.5 156 54.2 172 49.7 166 83.9 192 73.0 187 67.6 173 88.7 198 71.8 195 56.1 171 85.6 197 28.6 116 43.9 56 43.6 179 38.1 176 57.0 155 53.2 183
NL-TV-NCC [25]167.0 37.1 142 35.7 169 48.0 59 27.8 117 35.0 150 37.6 85 31.0 171 35.4 169 40.0 147 56.0 184 54.2 194 82.6 179 73.8 195 68.6 190 89.1 199 70.6 189 58.4 194 82.5 147 30.4 186 50.0 190 44.0 184 39.8 195 60.2 194 52.4 180
SLK [47]167.4 41.6 188 38.7 189 54.4 189 33.0 184 38.3 183 45.5 181 33.3 188 38.6 190 42.8 172 57.8 190 51.8 185 83.5 190 72.1 58 67.3 150 86.5 30 70.1 172 55.8 162 82.7 159 30.0 182 51.4 194 43.0 164 38.2 179 57.5 167 51.5 171
Learning Flow [11]170.2 37.7 158 37.0 179 49.2 144 29.9 165 37.5 176 39.7 129 31.5 181 36.3 175 40.7 159 55.4 181 51.6 184 82.6 179 72.8 183 67.8 184 87.8 187 69.6 160 55.7 160 82.8 165 29.3 164 48.4 177 42.7 149 39.2 191 59.8 191 51.2 165
GroupFlow [9]170.4 40.3 185 40.1 191 51.3 178 31.5 175 38.9 188 42.6 163 33.5 190 39.5 192 43.8 174 54.7 180 52.3 188 81.0 139 73.2 189 68.6 190 87.6 175 70.4 187 57.3 187 83.0 170 29.3 164 48.1 176 42.5 95 37.9 170 58.2 180 50.1 63
FFV1MT [104]170.4 39.6 180 40.8 193 50.3 167 34.8 190 38.8 186 46.6 185 36.5 195 45.8 196 44.6 178 56.2 186 49.0 153 81.7 166 72.5 161 67.4 162 87.4 78 70.2 180 54.7 110 83.2 174 30.1 184 49.3 188 42.8 155 38.6 186 57.6 169 51.3 167
Heeger++ [102]171.5 40.6 187 42.5 195 50.4 168 33.5 186 38.2 181 43.0 166 37.6 196 48.1 197 44.9 179 56.2 186 49.0 153 81.7 166 73.4 192 68.9 195 87.5 140 70.1 172 56.0 167 82.8 165 30.3 185 49.1 186 42.6 135 37.7 164 56.9 149 50.3 107
2bit-BM-tele [96]171.9 37.9 162 34.7 153 50.1 166 30.1 168 36.4 172 41.7 159 30.2 150 32.2 94 41.3 166 54.3 175 49.4 161 84.1 194 73.3 191 68.1 187 88.3 193 72.2 197 57.2 186 85.5 196 30.4 186 52.7 197 44.4 186 38.2 179 56.3 121 54.1 187
H+S_RVC [176]173.5 39.9 183 38.4 187 51.3 178 33.9 188 38.8 186 44.8 176 35.6 193 44.0 195 45.8 182 59.3 194 49.8 168 82.3 178 72.5 161 67.8 184 87.1 35 70.1 172 55.6 158 82.7 159 30.6 188 49.8 189 43.4 173 39.8 195 57.4 166 51.9 175
FOLKI [16]177.2 44.6 194 40.4 192 58.4 195 35.7 192 42.3 196 47.3 187 33.3 188 40.7 193 44.9 179 59.4 196 53.6 192 86.5 197 72.5 161 67.6 173 87.3 52 70.1 172 55.8 162 83.2 174 29.4 166 48.4 177 43.1 168 38.7 187 58.5 184 52.0 176
Pyramid LK [2]182.7 46.1 196 38.9 190 61.0 197 36.7 195 40.4 193 50.9 192 39.9 197 36.6 177 49.4 191 64.1 198 61.2 198 87.7 198 73.1 188 68.6 190 87.4 78 70.1 172 56.1 171 83.0 170 29.6 175 50.7 192 43.2 170 39.2 191 61.2 195 51.5 171
Adaptive flow [45]183.0 43.8 192 38.2 185 56.5 192 35.8 193 40.5 194 50.2 191 31.4 177 34.5 151 42.5 171 56.5 189 50.9 180 83.9 192 73.5 193 68.7 194 88.1 191 70.2 180 57.4 190 82.9 168 29.4 166 47.5 169 43.6 179 39.0 189 59.1 189 52.0 176
PGAM+LK [55]184.5 42.5 191 41.4 194 54.6 190 33.7 187 40.1 192 45.8 183 33.9 191 41.1 194 43.3 173 59.3 194 55.2 195 85.5 196 72.8 183 68.1 187 87.5 140 70.8 192 56.9 182 83.6 183 29.6 175 50.0 190 43.0 164 38.7 187 58.6 187 52.1 179
HCIC-L [97]188.5 49.1 198 42.6 196 63.0 198 35.8 193 39.4 191 52.5 194 34.6 192 37.7 185 43.9 175 58.0 191 53.9 193 81.7 166 74.0 196 69.3 196 88.5 196 71.7 194 60.5 197 83.2 174 29.5 171 47.5 169 43.9 183 40.5 197 62.9 197 52.8 181
Periodicity [79]193.9 44.4 193 43.3 197 56.9 193 42.8 198 43.4 197 56.2 197 40.9 198 49.1 198 49.5 192 58.9 193 58.6 197 84.9 195 74.4 197 70.2 197 88.0 189 73.1 198 57.9 193 86.2 198 30.0 182 51.5 195 43.5 175 41.8 198 63.4 198 53.7 186
AVG_FLOW_ROB [137]196.1 76.9 199 76.7 199 78.2 199 71.8 199 68.8 199 76.4 199 64.0 199 60.2 199 65.8 199 82.9 199 80.9 199 91.0 199 80.9 199 79.6 199 87.5 140 83.9 199 84.0 199 86.6 199 53.7 199 65.3 199 47.7 188 62.3 199 71.1 199 70.4 199
Move the mouse over the numbers in the table to see the corresponding images. Click to compare with the ground truth.

References

Methodtime*framescolor Reference and notes
[1] 2D-CLG 844 2 gray The 2D-CLG method by Bruhn et al. as implemented by Stefan Roth. [A. Bruhn, J. Weickert, and C. Schnörr. Lucas/Kanade meets Horn/Schunck: combining local and global optic flow methods. IJCV 63(3), 2005.] Parameters were set to match the published performance on Yosemite sequence, which may not be optimal for other sequences.
[2] Pyramid LK 12 2 color A modification of Bouguet's pyramidal implementation of Lucas-Kanade.
[3] Horn & Schunck 49 2 gray A modern Matlab implementation of the Horn & Schunck method by Deqing Sun. Parameters set to optimize AAE on all training data.
[4] Black & Anandan 328 2 gray A modern Matlab implementation of the Black & Anandan method by Deqing Sun.
[5] Brox et al. 18 2 color T. Brox, A. Bruhn, N. Papenberg, and J. Weickert. High accuracy optical flow estimation based on a theory for warping. ECCV 2004. (Improved using separate robust functions as proposed in A. Bruhn and J. Weickert, Towards ultimate motion estimation, ICCV 2005; improved by training on the training set.)
[6] Fusion 2,666 2 color V. Lempitsky, S. Roth, and C. Rother. Discrete-continuous optimization for optical flow estimation. CVPR 2008.
[7] Dynamic MRF 366 2 gray B. Glocker, N. Paragios, N. Komodakis, G. Tziritas, and N. Navab. Optical flow estimation with uncertainties through dynamic MRFs. CVPR 2008. (Method improved since publication.)
[8] Second-order prior 14 2 gray W. Trobin, T. Pock, D. Cremers, and H. Bischof. An unbiased second-order prior for high-accuracy motion estimation. DAGM 2008. (Method improved since publication; for details see W. Trobin, Ph.D. thesis, 2009.)
[9] GroupFlow 600 2 gray X. Ren. Local Grouping for Optical Flow. CVPR 2008.
[10] SegOF 60 2 color L. Xu, J. Chen, and J. Jia. Segmentation based variational model for accurate optical flow estimation. ECCV 2008. Code available.
[11] Learning Flow 825 2 gray D. Sun, S. Roth, J.P. Lewis, and M. Black. Learning optical flow (SRF-LFC). ECCV 2008.
[12] CBF 69 2 color W. Trobin, T. Pock, D. Cremers, and H. Bischof. Continuous energy minimization via repeated binary fusion. ECCV 2008. (Method improved since publication; for details see W. Trobin, Ph.D. thesis, 2009.)
[13] SPSA-learn 200 2 color Y. Li and D. Huttenlocher. Learning for optical flow using stochastic optimization. ECCV 2008.
[14] GraphCuts 1,200 2 color T. Cooke. Two applications of graph-cuts to image processing. DICTA 2008.
[15] F-TV-L1 8 2 gray A. Wedel, T. Pock, J. Braun, U. Franke, and D. Cremers. Duality TV-L1 flow with fundamental matrix prior. IVCNZ 2008.
[16] FOLKI 1.4 2 gray G. Le Besnerais and F. Champagnat. Dense optical flow by iterative local window registration. ICIP 2005.
[17] TV-L1-improved 2.9 2 gray A. Wedel, T. Pock, C. Zach, H. Bischof, and D. Cremers. An improved algorithm for TV-L1 optical flow computation. Proceedings of the Dagstuhl Visual Motion Analysis Workshop 2008. Code at GPU4Vision.
[18] DPOF 287 2 color C. Lei and Y.-H. Yang. Optical flow estimation on coarse-to-fine region-trees using discrete optimization. ICCV 2009. (Method improved since publication.)
[19] Filter Flow 34,000 2 color S. Seitz and S. Baker. Filter flow. ICCV 2009.
[20] Adaptive 9.2 2 gray A. Wedel, D. Cremers, T. Pock, and H. Bischof. Structure- and motion-adaptive regularization for high accuracy optic flow. ICCV 2009.
[21] Complementary OF 44 2 color H. Zimmer, A. Bruhn, J. Weickert, L. Valgaerts, A. Salgado, B. Rosenhahn, and H.-P. Seidel. Complementary optic flow. EMMCVPR 2009.
[22] Aniso. Huber-L1 2 2 gray M. Werlberger, W. Trobin, T. Pock, A. Wedel, D. Cremers, and H. Bischof. Anisotropic Huber-L1 optical flow. BMVC 2009. Code at GPU4Vision.
[23] Rannacher 0.12 2 gray J. Rannacher. Realtime 3D motion estimation on graphics hardware. Bachelor thesis, Heidelberg University, 2009.
[24] TI-DOFE 260 2 gray C. Cassisa, S. Simoens, and V. Prinet. Two-frame optical flow formulation in an unwarped multiresolution scheme. CIARP 2009.
[25] NL-TV-NCC 20 2 color M. Werlberger, T. Pock, and H. Bischof. Motion estimation with non-local total variation regularization. CVPR 2010.
[26] MDP-Flow 188 2 color L. Xu, J. Jia, and Y. Matsushita. Motion detail preserving optical flow estimation. CVPR 2010.
[27] ACK-Prior 5872 2 color K. Lee, D. Kwon, I. Yun, and S. Lee. Optical flow estimation with adaptive convolution kernel prior on discrete framework. CVPR 2010.
[28] LDOF 122 2 color T. Brox and J. Malik. Large displacement optical flow: descriptor matching in variational motion estimation. PAMI 33(3):500-513, 2011.
[29] p-harmonic 565 2 gray J. Gai and R. Stevenson. Optical flow estimation with p-harmonic regularization. ICIP 2010.
[30] TriangleFlow 4200 2 gray B. Glocker, H. Heibel, N. Navab, P. Kohli, and C. Rother. TriangleFlow: Optical flow with triangulation-based higher-order likelihoods. ECCV 2010.
[31] Classic+NL 972 2 color D. Sun, S. Roth, and M. Black. Secrets of optical flow estimation and their principles. CVPR 2010. Matlab code.
[32] Classic++ 486 2 gray A modern implementation of the classical formulation descended from Horn & Schunck and Black & Anandan; see D. Sun, S. Roth, and M. Black, Secrets of optical flow estimation and their principles, CVPR 2010.
[33] Nguyen 33 2 gray D. Nguyen. Tuning optical flow estimation with image-driven functions. ICRA 2011.
[34] Modified CLG 133 2 gray R. Fezzani, F. Champagnat, and G. Le Besnerais. Combined local global method for optic flow computation. EUSIPCO 2010.
[35] ComplOF-FED-GPU 0.97 2 color P. Gwosdek, H. Zimmer, S. Grewenig, A. Bruhn, and J. Weickert. A highly efficient GPU implementation for variational optic flow based on the Euler-Lagrange framework. CVGPU Workshop 2010.
[36] Ad-TV-NDC 35 2 gray M. Nawaz. Motion estimation with adaptive regularization and neighborhood dependent constraint. DICTA 2010.
[37] Layers++ 18206 2 color D. Sun, E. Sudderth, and M. Black. Layered image motion with explicit occlusions, temporal consistency, and depth ordering. NIPS 2010.
[38] OFH 620 3 color H. Zimmer, A. Bruhn, J. Weickert. Optic flow in harmony. IJCV 93(3) 2011.
[39] LSM 1615 2 color K. Jia, X. Wang, and X. Tang. Optical flow estimation using learned sparse model. ICCV 2011.
[40] CostFilter 55 2 color C. Rhemann, A. Hosni, M. Bleyer, C. Rother, and M. Gelautz. Fast cost-volume filtering for visual correspondence and beyond. CVPR 2011.
[41] Bartels 0.15 2 gray C. Bartels and G. de Haan. Smoothness constraints in recursive search motion estimation for picture rate conversion. IEEE TCSVT 2010. Version improved since publication: mapped on GPU.
[42] Shiralkar 600 2 gray M. Shiralkar and R. Schalkoff. A self organization-based optical flow estimator with GPU implementation. MVA 23(6):1229-1242.
[43] HBpMotionGpu 1000 5 gray S. Grauer-Gray and C. Kambhamettu. Hierarchical belief propagation to reduce search space using CUDA for stereo and motion estimation. WACV 2009. (Method improved since publication.)
[44] StereoFlow 7200 2 color G. Rosman, S. Shem-Tov, D. Bitton, T. Nir, G. Adiv, R. Kimmel, A. Feuer, and A. Bruckstein. Over-parameterized optical flow using a stereoscopic constraint. SSVM 2011:761-772.
[45] Adaptive flow 121 2 gray Tarik Arici and Vural Aksakalli. Energy minimization based motion estimation using adaptive smoothness priors. VISAPP 2012.
[46] TC-Flow 2500 5 color S. Volz, A. Bruhn, L. Valgaerts, and H. Zimmer. Modeling temporal coherence for optical flow. ICCV 2011.
[47] SLK 300 2 gray T. Corpetti and E. Mémin. Stochastic uncertainty models for the luminance consistency assumption. IEEE TIP 2011.
[48] CLG-TV 29 2 gray M. Drulea. Total variation regularization of local-global optical flow. ITSC 2011. Matlab code.
[49] SimpleFlow 1.7 2 color M. Tao, J. Bai, P. Kohli, S. Paris. SimpleFlow: a non-iterative, sublinear optical flow algorithm. EUROGRAPHICS 2012.
[50] IAOF 57 2 gray D. Nguyen. Improving motion estimation using image-driven functions and hybrid scheme. PSIVT 2011.
[51] IAOF2 56 2 gray Duc Dung Nguyen and Jae Wook Jeon. Enhancing accuracy and sharpness of motion field with adaptive scheme and occlusion-aware filter. IET Image Processing 7.2 (2013): 144-153.
[52] LocallyOriented 9541 2 gray Y.Niu, A. Dick, and M. Brooks. Locally oriented optical flow computation. To appear in TIP 2012.
[53] IROF-TV 261 2 color H. Rashwan, D. Puig, and M. Garcia. On improving the robustness of differential optical flow. ICCV 2011 Artemis workshop.
[54] Sparse Occlusion 2312 2 color Alper Ayvaci, Michalis Raptis, and Stefano Soatto. Sparse occlusion detection with optical flow. IJCV 97(3):322-338, 2012.
[55] PGAM+LK 0.37 2 gray A. Alba, E. Arce-Santana, and M. Rivera. Optical flow estimation with prior models obtained from phase correlation. ISVC 2010.
[56] Sparse-NonSparse 713 2 color Zhuoyuan Chen, Jiang Wang, and Ying Wu. Decomposing and regularizing sparse/non-sparse components for motion field estimation. CVPR 2012.
[57] nLayers 36150 4 color D. Sun, E. Sudderth, and M. Black. Layered segmentation and optical flow estimation over time. CVPR 2012.
[58] IROF++ 187 2 color H. Rashwan, D. Puig, and M. Garcia. Variational optical flow estimation based on stick tensor voting. IEEE TIP 2013.
[59] COFM 600 3 color M. Mozerov. Constrained optical flow estimation as a matching problem. IEEE TIP 2013.
[60] Efficient-NL 400 2 color P. Krähenbühl and V. Koltun. Efficient nonlocal regularization for optical flow. ECCV 2012.
[61] BlockOverlap 2 2 gray Michael Santoro, Ghassan AlRegib, and Yucel Altunbasak. Motion estimation using block overlap minimization. MMSP 2012.
[62] Ramp 1200 2 color A. Singh and N. Ahuja. Exploiting ramp structures for improving optical flow estimation. ICPR 2012.
[63] Occlusion-TV-L1 538 3 gray C. Ballester, L. Garrido, V. Lazcano, and V. Caselles. A TV-L1 optical flow method with occlusion detection. DAGM-OAGM 2012.
[64] TV-L1-MCT 90 2 color M. Mohamed and B. Mertsching. TV-L1 optical flow estimation with image details recovering based on modified census transform. ISVC 2012.
[65] Local-TV-L1 500 2 gray L. Raket. Local smoothness for global optical flow. ICIP 2012.
[66] ALD-Flow 61 2 color M. Stoll, A. Bruhn, and S. Volz. Adaptive integration of feature matches into variational optic flow methods. ACCV 2012.
[67] SIOF 234 2 color L. Xu, Z. Dai, and J. Jia. Scale invariant optical flow. ECCV 2012.
[68] MDP-Flow2 342 2 color L. Xu, J. Jia, and Y. Matsushita. Motion detail preserving optical flow estimation. PAMI 34(9):1744-1757, 2012. Code available.
[69] TCOF 1421 all gray J. Sanchez, A. Salgado, and N. Monzon. Optical flow estimation with consistent spatio-temporal coherence models. VISAPP 2013.
[70] LME 476 2 color W. Li, D. Cosker, M. Brown, and R. Tang. Optical flow estimation using Laplacian mesh energy. CVPR 2013.
[71] NN-field 362 2 color L. Chen, H. Jin, Z. Lin, S. Cohen, and Y. Wu. Large displacement optical flow from nearest neighbor fields. CVPR 2013.
[72] FESL 3310 2 color Weisheng Dong, Guangming Shi, Xiaocheng Hu, and Yi Ma. Nonlocal sparse and low-rank regularization for optical flow estimation. IEEE TIP 23(10):4527-4538, 2014.
[73] PMF 35 2 color J. Lu, H. Yang, D. Min, and M. Do. PatchMatch filter: efficient edge-aware filtering meets randomized search for fast correspondence field estimation. CVPR 2013.
[74] FC-2Layers-FF 2662 4 color D. Sun, J. Wulff, E. Sudderth, H. Pfister, and M. Black. A fully-connected layered model of foreground and background flow. CVPR 2013.
[75] NNF-Local 673 2 color Zhuoyuan Chen, Hailin Jin, Zhe Lin, Scott Cohen, and Ying Wu. Large displacement optical flow from nearest neighbor fields. CVPR 2013.
[76] Correlation Flow 290 2 color M. Drulea and S. Nedevschi. Motion estimation using the correlation transform. TIP 2013. Matlab code.
[77] TC/T-Flow 341 5 color M. Stoll, S. Volz, and A. Bruhn. Joint trilateral filtering for multiframe optical flow. ICIP 2013.
[78] OFLAF 1530 2 color T. Kim, H. Lee, and K. Lee. Optical flow via locally adaptive fusion of complementary data costs. ICCV 2013.
[79] Periodicity 8000 4 color Georgii Khachaturov, Silvia Gonzalez-Brambila, and Jesus Gonzalez-Trejo. Periodicity-based computation of optical flow. Computacion y Sistemas (CyS) 2014.
[80] SILK 572 2 gray Pascal Zille, Thomas Corpetti, Liang Shao, and Xu Chen. Observation model based on scale interactions for optical flow estimation. IEEE TIP 23(8):3281-3293, 2014.
[81] CRTflow 13 3 color O. Demetz, D. Hafner, and J. Weickert. The complete rank transform: a tool for accurate and morphologically invariant matching of structures. BMVC 2013.
[82] Classic+CPF 640 2 gray Zhigang Tu, Nico van der Aa, Coert Van Gemeren, and Remco Veltkamp. A combined post-filtering method to improve accuracy of variational optical flow estimation. Pattern Recognition 47(5):1926-1940, 2014.
[83] S2D-Matching 1200 2 color Marius Leordeanu, Andrei Zanfir, and Cristian Sminchisescu. Locally affine sparse-to-dense matching for motion and occlusion estimation. ICCV 2013.
[84] AGIF+OF 438 2 gray Zhigang Tu, Ronald Poppe, and Remco Veltkamp. Adaptive guided image filter for warping in variational optical flow computation. Signal Processing 127:253-265, 2016.
[85] DeepFlow 13 2 color P. Weinzaepfel, J. Revaud, Z. Harchaoui, and C. Schmid. DeepFlow: large displacement optical flow with deep matching. ICCV 2013.
[86] EPPM w/o HM 2.5 2 color L. Bao, Q. Yang, and H. Jin. Fast edge-preserving PatchMatch for large displacement optical flow. CVPR 2014.
[87] MLDP_OF 165 2 gray M. Mohamed, H. Rashwan, B. Mertsching, M. Garcia, and D. Puig. Illumination-robust optical flow approach using local directional pattern. IEEE TCSVT 24(9):1499-1508, 2014.
[88] RFlow 20 2 gray S. Ali, C. Daul, and W. Blondel. Robust and accurate optical flow estimation for weak texture and varying illumination condition: Application to cystoscopy. IPTA 2014.
[89] SRR-TVOF-NL 32 all color P. Pohl, M. Sirotenko, E. Tolstaya, and V. Bucha. Edge preserving motion estimation with occlusions correction for assisted 2D to 3D conversion. IS&T/SPIE Electronic Imaging 2014.
[90] 2DHMM-SAS 157 2 color M.-C. Shih, R. Shenoy, and K. Rose. A two-dimensional hidden Markov model with spatially-adaptive states with application of optical flow. ICIP 2014 submission.
[91] WLIF-Flow 700 2 color Z. Tu, R. Veltkamp, N. van der Aa, and C. Van Gemeren. Weighted local intensity fusion method for variational optical flow estimation. Submitted to TIP 2014.
[92] FMOF 215 2 color N. Jith, A. Ramakanth, and V. Babu. Optical flow estimation using approximate nearest neighbor field fusion. ICASSP 2014.
[93] TriFlow 150 2 color TriFlow. Optical flow with geometric occlusion estimation and fusion of multiple frames. ECCV 2014 submission 914.
[94] ComponentFusion 6.5 2 color Anonymous. Fast optical flow by component fusion. ECCV 2014 submission 941.
[95] AggregFlow 1642 2 color D. Fortun, P. Bouthemy, and C. Kervrann. Aggregation of local parametric candidates and exemplar-based occlusion handling for optical flow. Preprint arXiv:1407.5759.
[96] 2bit-BM-tele 124 2 gray R. Xu and D. Taubman. Robust dense block-based motion estimation using a two-bit transform on a Laplacian pyramid. ICIP 2013.
[97] HCIC-L 330 2 color Anonymous. Globally-optimal image correspondence using a hierarchical graphical model. NIPS 2014 submission 114.
[98] TF+OM 600 2 color R. Kennedy and C. Taylor. Optical flow with geometric occlusion estimation and fusion of multiple frames. EMMCVPR 2015.
[99] PH-Flow 800 2 color J. Yang and H. Li. Dense, accurate optical flow estimation with piecewise parametric model. CVPR 2015.
[100] EpicFlow 16 2 color J. Revaud, P. Weinzaepfel, Z. Harchaoui, and C. Schmid. EpicFlow: edge-preserving interpolation of correspondences for optical flow. CVPR 2015.
[101] NNF-EAC 380 2 color Anonymous. Variational method for joint optical flow estimation and edge-aware image restoration. CVPR 2015 submission 2336.
[102] Heeger++ 6600 5 gray Anonymous. A context aware biologically inspired algorithm for optical flow (updated results). CVPR 2015 submission 2238.
[103] HBM-GC 330 2 color A. Zheng and Y. Yuan. Motion estimation via hierarchical block matching and graph cut. Submitted to ICIP 2015.
[104] FFV1MT 358 5 gray F. Solari, M. Chessa, N. Medathati, and P. Kornprobst. What can we expect from a V1-MT feedforward architecture for optical flow estimation? Submitted to Signal Processing: Image Communication 2015.
[105] ROF-ND 4 2 color S. Ali, C. Daul, E. Galbrun, and W. Blondel. Illumination invariant large displacement optical flow using robust neighbourhood descriptors. Submitted to CVIU 2015.
[106] DeepFlow2 16 2 color J. Revaud, P. Weinzaepfel, Z. Harchaoui, and C. Schmid. Deep convolutional matching. Submitted to IJCV, 2015.
[107] HAST 2667 2 color Anonymous. Highly accurate optical flow estimation on superpixel tree. ICCV 2015 submission 2221.
[108] FlowFields 15 2 color C. Bailer, B. Taetz, and D. Stricker. Flow Fields: Dense unregularized correspondence fields for highly accurate large displacement optical flow estimation. ICCV 2015.
[109] SVFilterOh 1.56 2 color Anonymous. Fast estimation of large displacement optical flow using PatchMatch and dominant motion patterns. CVPR 2016 submission 1788.
[110] FlowNetS+ft+v 0.5 2 color Anonymous. Learning optical flow with convolutional neural networks. ICCV 2015 submission 235.
[111] CombBMOF 51 2 color M. Brüggemann, R. Kays, P. Springer, and O. Erdler. Combined block-matching and adaptive differential motion estimation in a hierarchical multi-scale framework. ICGIP 2014. (Method improved since publication.)
[112] PMMST 182 2 color F. Zhang, S. Xu, and X. Zhang. High accuracy correspondence field estimation via MST based patch matching. Submitted to TIP 2015.
[113] DF-Auto 70 2 color N. Monzon, A. Salgado, and J. Sanchez. Regularization strategies for discontinuity-preserving optical flow methods. Submitted to TIP 2015.
[114] CPM-Flow 3 2 color Anonymous. Efficient coarse-to-fine PatchMatch for large displacement optical flow. CVPR 2016 submission 241.
[115] CNN-flow-warp+ref 1.4 3 color D. Teney and M. Hebert. Learning to extract motion from videos in convolutional neural networks. ArXiv 1601.07532, 2016.
[116] Steered-L1 804 2 color Anonymous. Optical flow estimation via steered-L1 norm. Submitted to WSCG 2016.
[117] StereoOF-V1MT 343 2 gray Anonymous. Visual features for action-oriented tasks: a cortical-like model for disparity and optic flow computation. BMVC 2016 submission 132.
[118] PGM-C 5 2 color Y. Li. Pyramidal gradient matching for optical flow estimation. Submitted to PAMI 2016.
[119] RNLOD-Flow 1040 2 gray C. Zhang, Z. Chen, M. Wang, M. Li, and S. Jiang. Robust non-local TV-L1 optical flow estimation with occlusion detection. IEEE TIP 26(8):4055-4067, 2017.
[120] FlowNet2 0.091 2 color Anonymous. FlowNet 2.0: Evolution of optical flow estimation with deep networks. CVPR 2017 submission 900.
[121] S2F-IF 20 2 color Anonymous. S2F-IF: Slow-to-fast interpolator flow. CVPR 2017 submission 765.
[122] BriefMatch 0.068 2 gray G. Eilertsen, P.-E. Forssen, and J. Unger. Dense binary feature matching for real-time optical flow estimation. SCIA 2017 submission 62.
[123] OAR-Flow 60 2 color Anonymous. Order-adaptive regularisation for variational optical flow: global, local and in between. SSVM 2017 submission 20.
[124] AdaConv-v1 2.8 2 color Simon Niklaus, Long Mai, and Feng Liu. (Interpolation results only.) Video frame interpolation via adaptive convolution. CVPR 2017.
[125] SepConv-v1 0.2 2 color Simon Niklaus, Long Mai, and Feng Liu. (Interpolation results only.) Video frame interpolation via adaptive separable convolution. ICCV 2017.
[126] ProbFlowFields 37 2 color A. Wannenwetsch, M. Keuper, and S. Roth. ProbFlow: joint optical flow and uncertainty estimation. ICCV 2017.
[127] UnFlow 0.12 2 color Anonymous. UnFlow: Unsupervised learning of optical flow with a bidirectional census loss. Submitted to AAAI 2018.
[128] FlowFields+ 10.5 2 color C. Bailer, B. Taetz, and D. Stricker. Flow fields: Dense correspondence fields for highly accurate large displacement optical flow estimation. Submitted to PAMI 2017.
[129] IIOF-NLDP 150 2 color D.-H. Trinh, W. Blondel, and C. Daul. A general form of illumination-invariant descriptors in variational optical flow estimation. ICIP 2017.
[130] SuperSlomo 0.5 2 color Anonymous. (Interpolation results only.) Super SloMo: High quality estimation of multiple intermediate frames for video interpolation. CVPR 2018 submission 325.
[131] EPMNet 0.061 2 color Anonymous. EPM-convolution multilayer-network for optical flow estimation. ICME 2018 submission 1119.
[132] OFRF 90 2 color Tan Khoa Mai, Michele Gouiffes, and Samia Bouchafa. Optical flow refinement using iterative propagation under colour, proximity and flow reliability constraints. IET Image Processing 2020.
[133] 3DFlow 328 2 color J. Chen, Z. Cai, J. Lai, and X. Xie. A filtering based framework for optical flow estimation. IEEE TCSVT 2018.
[134] CtxSyn 0.07 2 color Simon Niklaus and Feng Liu. (Interpolation results only.) Context-aware synthesis for video frame interpolation. CVPR 2018.
[135] DMF_ROB 10 2 color ROB 2018 baseline submission, based on: P. Weinzaepfel, J. Revaud, Z. Harchaoui, and C. Schmid. DeepFlow: large displacement optical flow with deep matching. ICCV 2013.
[136] JOF 657 2 gray C. Zhang, L. Ge, Z. Chen, M. Li, W. Liu, and H. Chen. Refined TV-L1 optical flow estimation using joint filtering. Submitted to IEEE TMM, 2018.
[137] AVG_FLOW_ROB N/A 2 N/A Average flow field of ROB 2018 training set.
[138] LiteFlowNet 0.06 2 color T.-W. Hui, X. Tang, and C. C. Loy. LiteFlowNet: A lightweight convolutional neural network for optical flow estimation. CVPR 2018.
[139] AugFNG_ROB 0.10 all color Anonymous. FusionNet and AugmentedFlowNet: Selective proxy ground truth for training on unlabeled images. ECCV 2018 submission 2834.
[140] ResPWCR_ROB 0.2 2 color Anonymous. Learning optical flow with residual connections. ROB 2018 submission.
[141] FF++_ROB 17.43 2 color R. Schuster, C. Bailer, O. Wasenmueller, D. Stricker. FlowFields++: Accurate optical flow correspondences meet robust interpolation. ICIP 2018. Submitted to ROB 2018.
[142] ProFlow_ROB 76 3 color Anonymous. ProFlow: Learning to predict optical flow. BMVC 2018 submission 277.
[143] PWC-Net_RVC 0.069 2 color D. Sun, X. Yang, M.-Y. Liu, and J. Kautz. PWC-Net: CNNs for optical flow using pyramid, warping, and cost volume. CVPR 2018. Also RVC 2020 baseline submission.
[144] WOLF_ROB 0.02 2 color Anonymous. Reversed deep neural network for optical flow. ROB 2018 submission.
[145] LFNet_ROB 0.068 2 color Anonymous. Learning a flow network. ROB 2018 submission.
[146] WRT 9 2 color L. Mei, J. Lai, X. Xie, J. Zhu, and J. Chen. Illumination-invariance optical flow estimation using weighted regularization transform. Submitted to IEEE TCSVT 2018.
[147] EAI-Flow 2.1 2 color Anonymous. Hierarchical coherency sensitive hashing and interpolation with RANSAC for large displacement optical flow. CVIU 2018 submission 17-678.
[148] ContinualFlow_ROB 0.5 all color Michal Neoral, Jan Sochman, and Jiri Matas. Continual occlusions and optical flow estimation. ACCV 2018.
[149] CyclicGen 0.088 2 color Anonymous. (Interpolation results only.) Deep video frame interpolation using cyclic frame generation. AAAI 2019 submission 323.
[150] TOF-M 0.393 2 color Tianfan Xue, Baian Chen, Jiajun Wu, Donglai Wei, and William Freeman. Video enhancement with task-oriented flow. arXiv 1711.09078, 2017.
[151] MPRN 0.32 4 color Anonymous. (Interpolation results only.) Multi-frame pyramid refinement network for video frame interpolation. CVPR 2019 submission 1361.
[152] DAIN 0.13 2 color Wenbo Bao, Wei-Sheng Lai, Chao Ma, Xiaoyun Zhang, Zhiyong Gao, and Ming-Hsuan Yang. (Interpolation results only.) DAIN: Depth-aware video frame interpolation. CVPR 2019.
[153] FRUCnet 0.65 2 color Van Thang Nguyen, Kyujoong Lee, and Hyuk-Jae Lee. (Interpolation results only.) A stacked deep MEMC network for frame rate up conversion and its application to HEVC. Submitted to IEEE TCSVT 2019.
[154] OFRI 0.31 2 color Anonymous. (Interpolation results only.) Efficient video frame interpolation via optical flow refinement. CVPR 2019 submission 6743.
[155] CompactFlow_ROB 0.05 2 color Anonymous. CompactFlow: spatially shiftable window revisited. CVPR 2019 submission 1387.
[156] SegFlow 3.2 2 color Jun Chen, Zemin Cai, Jianhuang Lai, and Xiaohua Xie. Efficient segmentation-based PatchMatch for large displacement optical flow estimation. IEEE TCSVT 2018.
[157] HCFN 0.18 2 color Anonymous. Practical coarse-to-fine optical flow with deep networks. ICCV 2019 submission 116.
[158] FGME 0.23 2 color Bo Yan, Weimin Tan, Chuming Lin, and Liquan Shen. (Interpolation results only.) Fine-grained motion estimation for video frame interpolation. IEEE Transactions on Broadcasting, 2020.
[159] MS-PFT 0.44 2 color Xianhang Cheng and Zhenzhong Chen. (Interpolation results only.) A multi-scale position feature transform network for video frame interpolation. IEEE TCSVT 2020.
[160] MEMC-Net+ 0.12 2 color Wenbo Bao, Wei-Sheng Lai, Xiaoyun Zhang, Zhiyong Gao, and Ming-Hsuan Yang. (Interpolation results only.) MEMC-Net: Motion estimation and motion compensation driven neural network for video interpolation and enhancement. Submitted to PAMI 2018.
[161] ADC 0.01 2 color Anonymous. (Interpolation results only.) Learning spatial transform for video frame interpolation. ICCV 2019 submission 5424.
[162] DSepConv 0.3 2 color Xianhang Cheng and Zhenzhong Chen. (Interpolation results only.) Video frame interpolation via deformable separable convolution. AAAI 2020.
[163] MAF-net 0.3 2 color Mengshun Hu, Jing Xiao, Liang Liao, Zheng Wang, Chia-Wen Lin, Mi Wang, and Shinichi Satoh. Capturing small, fast-moving objects: Frame interpolation via recurrent motion enhancement. IEEE TCSVT 2021.
[164] STAR-Net 0.049 2 color Anonymous. (Interpolation results only.) Space-time-aware multiple resolution for video enhancement. CPVR 2020 submission 430.
[165] AdaCoF 0.03 2 color Hyeongmin Lee, Taeoh Kim, Tae-young Chung, Daehyun Pak, Yuseok Ban, and Sangyoun Lee. (Interpolation results only.) AdaCoF: Adaptive collaboration of flows for video frame interpolation. CVPR 2020. Code available.
[166] TC-GAN 0.13 2 color Anonymous. (Interpolation results only.) A temporal and contextual generative adversarial network for video frame interpolation. CVPR 2020 submission 111.
[167] FeFlow 0.52 2 color Shurui Gui, Chaoyue Wang, Qihua Chen, and Dacheng Tao. (Interpolation results only.) FeatureFlow: Robust video interpolation via structure-to-texture generation. CVPR 2020. Code available.
[168] DAI 0.23 2 color Anonymous. (Interpolation results only.) Deep animation inbetweening. CVPR 2020 submission 6404.
[169] SoftSplat 0.1 2 color Simon Niklaus and Feng Liu. (Interpolation results only.) Softmax splatting for video frame interpolation. CVPR 2020.
[170] STSR 5.35 2 color Anonymous. (Interpolation results only.) Spatial and temporal video super-resolution with a frequency domain loss. ECCV 2020 submission 2340.
[171] BMBC 0.77 2 color Anonymous. (Interpolation results only.) BMBC: Bilateral motion estimation with bilateral cost volume for video interpolation. ECCV 2020 submission 2095.
[172] GDCN 1.0 2 color Anonymous. (Interpolation results only.) Video interpolation via generalized deformable convolution. ECCV 2020 submission 4347.
[173] EDSC 0.56 2 color Xianhang Cheng and Zhenzhong Chen. (Interpolation results only.) Multiple video frame interpolation via enhanced deformable separable convolution. Submitted to PAMI 2020.
[174] CoT-AMFlow 0.04 2 color Anonymous. CoT-AMFlow: Adaptive modulation network with co-teaching strategy for unsupervised optical flow estimation. CoRL 2020 submission 36.
[175] TVL1_RVC 11.6 2 color RVC 2020 baseline submission by Toby Weed, based on: Javier Sanchez, Enric Meinhardt-Llopis, and Gabriele Facciolo. TV-L1 optical flow estimation. IPOL 3:137-150, 2013.
[176] H+S_RVC 44.7 2 color RVC 2020 baseline submission by Toby Weed, based on: Enric Meinhardt-Llopis, Javier Sanchez, and Daniel Kondermann. Horn-Schunck optical flow with a multi-scale strategy. IPOL 3:151–172, 2013.
[177] PRAFlow_RVC 0.34 2 color Zhexiong Wan, Yuxin Mao, and Yuchao Dai. Pyramid recurrent all-pairs flow. RVC 2020 submission.
[178] VCN_RVC 0.84 2 color Gengshan Yang and Deva Ramanan. Volumetric correspondence networks for optical flow. NeurIPS 2019. RVC 2020 submission.
[179] RAFT-TF_RVC 1.51 2 color Deqing Sun, Charles Herrmann, Varun Jampani, Mike Krainin, Forrester Cole, Austin Stone, Rico Jonschkowski, Ramin Zabih, William Freeman, and Ce Liu. A TensorFlow implementation of RAFT (Zachary Teed and Jia Deng. RAFT: Recurrent all-pairs field transforms for optical flow. ECCV 2020.) RVC 2020 submission.
[180] IRR-PWC_RVC 0.18 2 color Junhwa Hur and Stefan Roth. Iterative residual refinement for joint optical flow and occlusion estimation. CVPR 2019. RVC 2020 submission.
[181] C-RAFT_RVC 0.60 2 color Henrique Morimitsu and Xiangyang Ji. Classification RAFT. RVC 2020 submission.
[182] LSM_FLOW_RVC 0.2 2 color Chengzhou Tang, Lu Yuan, and Ping Tan. LSM: Learning subspace minimization for low-level vision. CVPR 2020. RVC 2020 submission.
[183] MV_VFI 0.23 2 color Zhenfang Wang, Yanjiang Wang, and Baodi Liu. (Interpolation results only.) Multi-view based video interpolation algorithm. ICASSP 2021 submission.
[184] DistillNet 0.12 2 color Anonymous. (Interpolation results only.) A teacher-student optical-flow distillation framework for video frame interpolation. CVPR 2021 submission 7534.
[185] SepConv++ 0.1 2 color Simon Niklaus, Long Mai, and Oliver Wang. (Interpolation results only.) Revisiting adaptive convolutions for video frame interpolation. WACV 2021.
[186] EAFI 0.18 2 color Anonymous. (Interpolation results only.) Error-aware spatial ensembles for video frame interpolation. ICCV 2021 submission 8020.
[187] UnDAF 0.04 2 color Anonymous. UnDAF: A general unsupervised domain adaptation framework for disparity, optical flow or scene flow estimation. CVPR 2021 submission 236.
[188] FLAVR 0.029 all color Anonymous. (Interpolation results only.) FLAVR frame interpolation. NeurIPS 2021 submission 1300.
[189] PBOFVI 150 2 color Zemin Cai, Jianhuang Lai, Xiaoxin Liao, and Xucong Chen. Physics-based optical flow under varying illumination. Submitted to IEEE TCSVT 2021.
[190] SoftsplatAug 0.17 2 color Anonymous. (Interpolation results only.) Transformation data augmentation for sports video frame interpolation. ICCV 2021 submission 3245.
* The "time" column lists the reported runtime in seconds on the "Urban" sequence. Note that these runtimes are not normalized by processor speed or type.