Optical flow evaluation results Statistics:     Average   SD   R2.5   R5.0   R10.0   A90   A95   A99  
Error type:   endpoint   angle   interpolation   normalized interpolation  
Show images: below table   above table        
A99
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
PMMST [114]10.2 11.2 10 21.1 6 2.71 2 13.8 8 19.7 7 3.65 22 10.3 1 19.2 7 2.71 1 16.8 4 30.8 22 7.53 30 41.1 5 51.1 7 10.0 17 24.6 2 43.0 6 4.93 12 34.2 10 70.9 12 4.04 8 28.8 14 45.4 24 3.42 9
MDP-Flow2 [68]10.7 11.0 3 20.7 5 2.71 2 13.9 10 19.9 9 3.46 2 10.3 1 20.3 14 3.00 2 16.7 3 30.0 14 7.35 7 41.0 4 50.7 4 10.1 29 27.1 35 44.9 19 4.97 19 33.6 3 70.1 7 3.92 3 29.2 18 47.0 35 3.42 9
NNF-Local [87]19.9 11.4 16 21.6 9 2.71 2 12.8 1 18.4 2 3.56 4 10.4 3 20.0 12 3.00 2 19.8 58 37.3 88 7.35 7 41.5 11 51.4 9 10.0 17 28.2 61 47.3 37 5.07 45 34.5 13 71.9 23 4.04 8 29.1 17 46.1 30 3.37 2
SepConv-v1 [127]20.1 9.68 1 19.1 1 2.52 1 15.4 34 20.1 11 5.26 99 11.0 10 16.7 1 3.87 93 20.4 71 26.8 1 9.59 115 41.9 12 52.5 16 9.00 2 24.7 3 42.4 3 4.69 1 30.7 1 67.4 1 3.92 3 24.7 1 35.8 1 3.32 1
NN-field [71]22.7 11.5 24 22.9 23 2.71 2 13.0 3 18.6 3 3.42 1 12.3 76 19.7 8 3.00 2 21.1 78 39.8 102 7.44 17 41.4 9 51.4 9 10.0 17 27.5 43 46.4 28 4.97 19 33.8 6 71.0 14 4.04 8 29.3 20 46.2 31 3.37 2
NNF-EAC [103]26.4 11.5 24 21.7 10 3.11 71 14.5 21 21.0 23 3.70 24 12.3 76 22.6 35 3.00 2 17.7 16 32.4 41 7.55 36 43.2 38 55.1 42 10.1 29 25.1 6 43.8 9 4.90 6 34.0 8 70.5 9 4.08 37 29.4 22 47.5 40 3.42 9
DeepFlow2 [108]28.2 11.4 16 23.5 27 3.00 58 16.7 56 23.0 58 4.04 54 11.0 10 20.3 14 3.00 2 19.0 47 29.8 12 7.53 30 42.7 25 54.0 25 10.3 52 25.0 4 43.0 6 4.93 12 35.2 27 73.8 34 4.04 8 28.9 15 44.9 21 3.56 65
DeepFlow [86]28.3 11.3 14 24.2 40 3.00 58 16.6 55 23.0 58 4.32 67 11.0 10 20.3 14 3.00 2 19.3 49 28.1 6 7.59 44 42.7 25 54.5 30 10.2 47 25.2 8 44.1 10 5.00 39 32.9 2 68.2 2 4.04 8 28.4 10 44.6 17 3.56 65
SuperFlow [81]29.4 11.0 3 22.1 17 3.11 71 17.1 61 22.7 50 4.69 80 11.7 45 18.7 2 3.37 65 18.7 40 27.4 3 7.70 64 41.3 7 51.2 8 9.98 13 26.3 22 46.9 31 4.80 3 34.7 16 76.0 45 4.08 37 28.1 8 41.5 5 3.42 9
PH-Flow [101]30.1 11.9 55 25.7 69 2.83 16 13.3 5 19.7 7 3.56 4 10.7 5 22.7 36 3.00 2 16.5 2 30.2 15 7.33 3 42.3 18 52.1 13 10.1 29 28.7 76 50.9 90 5.20 69 35.6 34 77.0 55 4.04 8 29.6 28 47.0 35 3.51 49
DF-Auto [115]30.2 10.9 2 19.2 2 3.11 71 17.2 64 23.4 66 4.43 72 10.4 3 20.6 21 3.00 2 18.1 27 29.7 9 7.55 36 41.4 9 52.1 13 10.0 17 26.2 17 47.2 34 4.97 19 35.2 27 79.3 66 4.08 37 29.6 28 44.7 18 3.56 65
CombBMOF [113]30.7 12.0 60 24.3 41 2.83 16 14.3 15 20.6 16 3.56 4 11.3 26 25.7 60 3.00 2 20.3 70 34.9 66 7.55 36 43.2 38 54.0 25 10.1 29 26.4 23 47.7 44 4.90 6 36.2 51 71.4 18 4.08 37 29.5 25 45.7 27 3.37 2
CBF [12]32.1 11.0 3 19.8 3 3.00 58 17.1 61 22.9 55 4.24 64 12.0 60 19.0 4 3.00 2 17.8 21 28.0 5 7.85 78 40.6 3 49.9 2 9.97 9 26.2 17 44.6 14 4.97 19 36.3 53 76.3 50 4.12 80 27.9 4 41.2 4 3.70 102
Aniso. Huber-L1 [22]33.0 11.4 16 21.7 10 3.11 71 19.7 99 24.7 99 4.55 76 12.0 60 19.7 8 3.11 60 18.4 32 29.8 12 7.55 36 42.5 21 54.4 29 9.98 13 25.2 8 42.2 2 4.83 4 35.6 34 71.5 19 4.04 8 27.9 4 42.0 7 3.56 65
LME [70]33.5 11.4 16 22.0 15 2.71 2 15.1 29 21.8 34 3.87 47 11.3 26 36.0 117 3.00 2 17.4 8 32.0 38 7.48 21 44.5 61 57.0 59 11.4 121 27.6 45 47.2 34 4.97 19 33.6 3 69.7 4 4.04 8 30.0 34 48.6 51 3.42 9
WLIF-Flow [93]34.6 11.5 24 22.1 17 2.83 16 15.2 30 21.6 31 3.79 38 11.3 26 26.4 68 3.00 2 17.4 8 30.3 17 7.59 44 42.5 21 53.5 21 10.4 61 29.0 82 51.1 93 5.29 85 34.8 19 69.7 4 4.04 8 30.0 34 48.4 47 3.46 34
CLG-TV [48]34.7 11.1 8 21.8 13 3.11 71 18.8 83 24.0 80 4.43 72 11.3 26 20.0 12 3.70 80 18.6 38 28.9 7 7.72 69 42.8 28 55.0 41 10.0 17 25.0 4 42.9 5 4.93 12 36.0 45 71.6 20 4.04 8 29.0 16 44.0 13 3.56 65
IROF++ [58]34.9 11.9 55 24.1 37 2.83 16 14.7 24 21.3 25 3.56 4 12.1 73 29.0 87 3.00 2 16.3 1 27.9 4 7.35 7 43.9 50 56.0 50 11.1 92 26.4 23 47.0 33 4.93 12 34.5 13 72.3 24 4.08 37 30.3 44 49.3 59 3.56 65
FMOF [94]36.4 12.2 80 24.5 50 2.94 46 14.0 11 20.0 10 3.56 4 12.3 76 27.7 75 3.00 2 19.8 58 35.4 71 7.70 64 42.4 19 52.1 13 10.1 29 28.1 57 49.1 60 4.93 12 34.6 15 72.7 28 3.87 1 30.2 40 47.6 43 3.42 9
Brox et al. [5]36.8 11.4 16 24.9 62 2.94 46 15.9 41 22.2 42 4.04 54 11.3 26 21.0 22 3.37 65 18.4 32 27.0 2 7.59 44 42.2 16 53.3 19 10.0 17 28.2 61 51.5 96 5.00 39 36.8 56 88.0 93 4.04 8 28.4 10 42.3 8 3.42 9
IROF-TV [53]37.5 11.7 38 24.7 57 3.00 58 15.5 37 22.0 40 3.70 24 11.0 10 23.7 46 3.00 2 17.3 6 31.3 26 7.57 43 43.8 48 56.0 50 11.2 99 27.6 45 48.4 51 4.97 19 35.9 43 74.5 41 4.08 37 28.0 6 42.6 9 3.56 65
ALD-Flow [66]38.3 12.0 60 28.4 93 3.11 71 16.3 48 22.8 52 3.83 42 11.0 10 21.7 28 3.00 2 17.9 24 33.6 54 7.39 12 43.4 43 54.6 34 10.8 79 25.8 12 44.8 18 5.00 39 34.1 9 70.4 8 4.04 8 31.9 73 50.3 67 3.46 34
nLayers [57]39.2 11.8 46 22.9 23 2.83 16 14.1 12 20.4 13 3.56 4 11.0 10 19.7 8 3.00 2 18.3 30 34.2 62 7.39 12 46.7 110 60.1 106 11.0 86 27.9 51 50.1 67 5.20 69 35.5 32 72.6 27 4.08 37 30.8 49 49.3 59 3.42 9
LDOF [28]40.4 11.4 16 22.5 20 3.56 112 16.1 43 21.4 30 6.35 116 12.0 60 20.3 14 3.70 80 19.0 47 29.7 9 7.94 83 41.2 6 50.9 5 10.1 29 26.8 28 50.2 69 4.90 6 34.8 19 80.2 70 4.08 37 29.4 22 44.5 15 3.46 34
p-harmonic [29]40.5 11.4 16 23.5 27 2.83 16 19.1 87 24.3 88 4.80 85 11.3 26 22.0 30 3.70 80 20.9 76 31.7 32 7.62 50 42.6 24 54.2 27 10.1 29 25.7 11 43.5 8 5.07 45 36.1 49 71.8 21 4.08 37 29.6 28 46.5 32 3.51 49
Layers++ [37]40.8 11.4 16 21.7 10 2.94 46 12.8 1 18.2 1 3.46 2 11.0 10 26.7 71 3.00 2 17.7 16 32.9 46 7.53 30 46.6 108 60.9 117 10.6 72 30.9 113 60.2 118 5.00 39 34.9 25 72.7 28 3.87 1 29.9 33 47.5 40 3.46 34
MDP-Flow [26]41.1 11.2 10 21.2 7 2.71 2 14.2 13 20.5 15 3.70 24 10.7 5 19.0 4 3.00 2 19.7 56 32.4 41 7.70 64 44.2 53 57.0 59 11.2 99 30.0 102 51.4 95 5.51 113 36.1 49 72.9 31 4.08 37 30.8 49 48.4 47 3.42 9
Second-order prior [8]41.2 11.3 14 22.0 15 3.11 71 19.0 86 24.2 86 4.32 67 13.3 87 27.7 75 3.70 80 18.8 43 31.6 31 7.51 25 42.9 30 54.7 36 10.0 17 26.2 17 45.0 20 4.97 19 35.6 34 71.2 15 4.04 8 29.5 25 45.4 24 3.56 65
SIOF [67]41.8 11.7 38 23.1 25 3.11 71 19.4 94 24.8 102 4.76 82 11.3 26 25.7 60 3.11 60 18.4 32 31.4 28 8.04 88 40.3 2 50.3 3 9.95 7 25.8 12 45.3 22 4.97 19 33.9 7 71.2 15 4.08 37 30.0 34 47.4 37 3.70 102
Local-TV-L1 [65]42.5 11.2 10 21.5 8 3.56 112 19.6 97 24.4 91 5.57 105 11.0 10 19.1 6 3.00 2 18.3 30 30.4 20 7.87 81 42.8 28 54.5 30 10.2 47 26.2 17 44.7 15 5.45 101 34.2 10 76.1 48 4.08 37 28.0 6 42.8 11 3.65 99
COFM [59]43.2 11.8 46 24.3 41 2.94 46 14.5 21 20.9 20 3.65 22 11.0 10 26.4 68 3.00 2 17.4 8 32.3 39 7.35 7 44.2 53 55.1 42 10.1 29 30.0 102 54.4 112 5.20 69 35.8 39 79.3 66 4.08 37 31.2 55 48.8 54 3.51 49
ProbFlowFields [128]44.5 11.6 30 25.4 66 2.83 16 14.4 17 21.1 24 3.56 4 10.7 5 23.7 46 3.00 2 18.4 32 33.4 51 7.59 44 46.2 98 59.2 93 11.2 99 28.5 73 50.7 84 5.32 89 34.7 16 76.9 53 4.08 37 29.4 22 46.5 32 3.46 34
FlowFields [110]44.8 11.8 46 25.6 68 2.83 16 14.4 17 20.9 20 3.56 4 11.3 26 24.3 52 3.00 2 20.0 64 38.1 93 7.51 25 43.6 45 54.5 30 11.0 86 28.2 61 50.7 84 5.16 64 34.8 19 75.1 44 4.04 8 32.0 78 52.0 90 3.46 34
TV-L1-MCT [64]46.8 12.4 96 24.7 57 2.83 16 16.4 49 23.1 60 3.83 42 11.9 59 32.7 108 3.00 2 17.6 13 31.7 32 7.53 30 47.0 118 61.2 118 11.0 86 25.5 10 44.7 15 4.97 19 36.0 45 80.7 75 4.04 8 28.4 10 44.8 20 3.46 34
BlockOverlap [61]47.1 11.1 8 20.1 4 3.56 112 19.3 91 23.7 73 6.16 112 11.3 26 20.4 19 3.70 80 18.4 32 29.6 8 8.72 107 43.1 35 54.5 30 10.2 47 27.4 40 48.6 54 5.35 96 34.8 19 72.8 30 4.08 37 27.2 3 40.9 3 3.56 65
Sparse-NonSparse [56]48.2 12.0 60 24.3 41 2.83 16 15.0 27 21.3 25 3.56 4 11.7 45 29.0 87 3.00 2 17.6 13 29.7 9 7.39 12 45.7 84 59.3 94 11.0 86 28.8 77 48.7 56 5.07 45 38.6 82 90.1 104 4.04 8 32.4 85 51.8 86 3.42 9
HAST [109]48.4 11.7 38 23.6 29 2.94 46 13.8 8 19.6 6 3.56 4 12.0 60 31.7 104 3.00 2 17.8 21 31.7 32 7.14 1 45.3 77 57.0 59 9.97 9 33.7 122 62.8 125 5.10 61 38.4 77 88.4 95 4.04 8 33.0 94 51.0 74 3.42 9
Modified CLG [34]49.0 11.0 3 21.9 14 3.11 71 19.6 97 23.9 77 5.94 110 12.4 80 26.3 66 3.87 93 19.8 58 30.8 22 8.12 93 42.1 14 52.9 17 10.1 29 27.0 31 48.1 49 5.23 75 34.7 16 70.8 11 4.08 37 29.5 25 45.3 23 3.56 65
OAR-Flow [125]49.1 12.0 60 24.9 62 3.00 58 16.4 49 22.4 45 4.08 59 11.0 10 20.5 20 3.00 2 17.4 8 33.6 54 7.33 3 46.2 98 60.0 105 11.3 114 27.0 31 47.6 41 5.23 75 37.6 66 74.0 36 4.08 37 31.0 54 49.2 57 3.46 34
CPM-Flow [116]49.7 11.8 46 27.3 83 2.83 16 14.4 17 20.4 13 3.70 24 11.7 45 24.0 49 3.00 2 21.4 85 40.1 104 7.77 73 45.5 81 58.1 75 11.2 99 26.6 27 48.0 47 5.07 45 36.0 45 72.3 24 4.04 8 30.9 52 50.4 68 3.56 65
AdaConv-v1 [126]50.1 15.0 121 28.2 92 3.70 116 17.6 69 20.7 18 7.68 124 17.4 104 22.0 30 7.00 118 27.5 114 33.7 58 17.0 125 39.9 1 49.8 1 8.19 1 23.8 1 39.5 1 4.76 2 34.2 10 68.5 3 4.12 80 26.9 2 39.5 2 3.42 9
2DHMM-SAS [92]50.5 12.2 80 24.5 50 2.83 16 17.9 72 24.1 83 3.87 47 12.0 60 28.7 84 3.00 2 17.3 6 31.4 28 7.51 25 45.1 73 58.2 79 11.2 99 27.9 51 49.0 58 4.83 4 37.0 58 76.1 48 4.08 37 31.9 73 50.5 69 3.42 9
F-TV-L1 [15]50.5 12.0 60 26.5 79 3.56 112 19.2 89 24.7 99 4.83 87 11.7 45 21.5 26 4.00 95 19.3 49 32.7 44 7.68 59 43.1 35 55.3 45 9.83 3 25.1 6 42.8 4 5.07 45 34.8 19 74.0 36 4.16 88 28.5 13 42.7 10 3.56 65
ComponentFusion [96]51.1 12.0 60 29.6 100 2.71 2 14.5 21 21.3 25 3.56 4 11.0 10 22.0 30 3.00 2 18.8 43 36.2 83 7.33 3 45.5 81 58.2 79 10.7 76 27.2 36 46.3 26 4.97 19 40.5 101 93.3 112 4.12 80 34.4 108 58.3 116 3.42 9
AGIF+OF [85]51.1 12.2 80 24.3 41 2.71 2 15.2 30 21.8 34 3.70 24 11.7 45 27.7 75 3.00 2 18.0 25 33.0 48 7.55 36 45.8 88 58.8 90 11.2 99 30.0 102 53.4 107 5.07 45 35.4 29 74.8 43 3.92 3 32.2 81 52.6 96 3.37 2
TC/T-Flow [76]52.4 12.4 96 26.4 78 2.83 16 16.5 53 23.1 60 3.83 42 11.0 10 22.4 34 3.00 2 18.9 45 34.5 63 7.33 3 45.5 81 58.1 75 11.4 121 27.3 39 47.6 41 4.93 12 41.1 103 80.4 73 4.20 94 30.9 52 49.7 63 3.37 2
DPOF [18]52.5 12.3 89 29.4 99 3.11 71 13.3 5 19.1 5 3.56 4 15.7 95 25.2 58 3.70 80 19.4 51 37.5 90 7.59 44 43.1 35 54.6 34 10.0 17 29.1 87 49.7 63 4.90 6 36.6 54 77.0 55 4.08 37 31.5 63 50.5 69 3.51 49
Ad-TV-NDC [36]54.0 12.2 80 22.5 20 4.32 123 20.6 118 24.8 102 5.80 106 11.7 45 21.6 27 3.37 65 21.6 86 31.8 35 8.04 88 42.5 21 53.4 20 9.97 9 26.4 23 47.6 41 5.16 64 36.8 56 70.9 12 4.08 37 28.3 9 41.8 6 3.70 102
PGM-C [120]54.5 11.8 46 27.3 83 2.83 16 14.4 17 20.7 18 3.70 24 12.3 76 23.0 40 3.00 2 20.6 74 42.3 111 7.62 50 45.8 88 59.5 99 11.2 99 27.2 36 47.4 38 4.97 19 37.1 60 79.2 64 4.04 8 32.4 85 55.0 107 3.51 49
Ramp [62]54.9 12.0 60 24.6 53 2.94 46 14.8 25 21.3 25 3.70 24 11.7 45 29.4 93 3.00 2 16.9 5 30.3 17 7.39 12 45.4 79 58.5 82 11.0 86 30.2 107 50.9 90 5.23 75 39.8 94 89.6 101 4.04 8 32.4 85 52.5 94 3.42 9
TF+OM [100]55.0 11.6 30 30.1 105 3.11 71 15.0 27 21.6 31 4.04 54 11.7 45 24.0 49 3.00 2 21.3 82 39.0 101 7.68 59 44.3 55 56.7 57 10.3 52 28.8 77 50.4 76 5.07 45 37.7 67 83.5 87 4.08 37 29.2 18 46.0 28 3.56 65
ComplOF-FED-GPU [35]55.2 12.0 60 27.9 89 2.94 46 15.7 40 22.2 42 3.79 38 16.0 96 21.4 24 3.70 80 18.4 32 33.6 54 7.48 21 44.9 70 57.7 67 10.7 76 27.4 40 45.9 23 5.00 39 36.6 54 78.7 63 4.08 37 32.6 92 52.3 92 3.51 49
PMF [73]55.2 12.2 80 25.9 72 2.71 2 15.4 34 21.8 34 3.56 4 12.7 82 35.7 115 3.00 2 20.2 68 35.9 77 7.51 25 44.4 59 54.9 40 10.1 29 28.4 66 50.5 79 5.32 89 37.9 71 81.1 77 4.04 8 34.2 107 54.1 102 3.37 2
AggregFlow [97]55.4 13.7 114 37.1 119 3.11 71 16.2 46 22.6 48 4.04 54 11.0 10 23.3 45 3.00 2 21.8 88 40.7 107 7.66 57 43.2 38 53.5 21 10.3 52 27.0 31 46.0 24 5.00 39 38.0 72 82.4 84 4.08 37 31.9 73 51.9 89 3.42 9
S2F-IF [123]56.4 12.1 73 29.8 102 2.71 2 14.2 13 20.6 16 3.56 4 11.3 26 26.3 66 3.00 2 20.2 68 40.1 104 7.53 30 45.9 92 58.7 89 11.3 114 28.4 66 50.7 84 5.20 69 35.7 37 76.0 45 4.08 37 32.3 83 53.1 97 3.46 34
OFLAF [77]56.5 11.7 38 24.5 50 2.71 2 13.6 7 20.3 12 3.56 4 11.0 10 23.0 40 3.00 2 17.6 13 31.3 26 7.39 12 47.3 120 61.7 121 11.2 99 29.6 97 51.9 102 5.32 89 41.8 108 95.6 117 4.16 88 33.6 100 52.1 91 3.42 9
Classic++ [32]57.0 11.6 30 23.7 30 3.11 71 17.8 71 24.4 91 4.08 59 11.7 45 20.3 14 3.37 65 20.1 66 33.8 59 7.62 50 44.7 64 57.8 70 10.0 17 28.0 54 49.7 63 5.35 96 37.4 64 81.4 79 4.08 37 30.7 48 49.5 61 3.56 65
Classic+NL [31]57.2 12.1 73 24.3 41 3.00 58 15.3 33 21.8 34 3.70 24 11.7 45 29.4 93 3.00 2 17.4 8 31.4 28 7.53 30 45.7 84 59.4 96 10.8 79 29.0 82 49.8 66 5.10 61 39.6 91 90.4 106 4.08 37 32.2 81 51.8 86 3.46 34
FlowNetS+ft+v [112]58.0 11.5 24 23.7 30 3.46 110 19.9 104 24.6 97 7.87 126 12.0 60 21.1 23 3.37 65 19.5 53 30.6 21 8.91 110 43.7 46 56.6 56 11.2 99 26.0 14 44.5 13 4.97 19 38.6 82 87.8 91 4.08 37 30.0 34 46.0 28 3.51 49
FC-2Layers-FF [74]58.9 12.1 73 26.0 76 2.83 16 13.0 3 18.7 4 3.56 4 11.4 42 25.7 60 3.00 2 17.8 21 33.5 53 7.48 21 46.5 104 60.3 111 11.2 99 30.4 110 52.3 106 5.32 89 39.8 94 90.0 103 4.08 37 31.8 69 51.6 82 3.46 34
LSM [39]59.0 12.3 89 24.7 57 2.83 16 15.4 34 21.9 38 3.56 4 12.0 60 30.3 99 3.00 2 18.7 40 33.2 50 7.44 17 46.1 96 59.4 96 11.1 92 29.3 90 51.9 102 5.07 45 39.2 87 91.0 108 4.04 8 32.3 83 52.5 94 3.42 9
MLDP_OF [89]59.0 11.9 55 24.7 57 2.83 16 17.4 67 23.8 75 3.87 47 10.7 5 24.6 55 3.00 2 20.5 73 33.6 54 8.35 99 44.1 51 56.5 54 10.1 29 29.3 90 50.5 79 5.57 114 35.8 39 73.4 33 4.20 94 31.2 55 50.6 72 3.70 102
TCOF [69]60.1 12.0 60 24.7 57 2.83 16 20.3 113 26.4 126 5.07 92 11.1 25 29.0 87 3.00 2 17.7 16 32.4 41 7.68 59 43.2 38 55.5 46 9.97 9 28.8 77 46.3 26 5.07 45 41.2 106 94.9 115 4.08 37 31.8 69 51.3 78 3.70 102
CRTflow [80]60.1 11.7 38 24.4 48 3.32 99 19.5 96 24.9 105 4.51 74 12.0 60 22.7 36 4.00 95 18.1 27 30.3 17 7.68 59 45.0 71 58.1 75 11.3 114 26.0 14 45.1 21 4.97 19 37.7 67 87.9 92 4.08 37 30.8 49 50.2 64 3.56 65
RNLOD-Flow [121]60.1 11.8 46 24.6 53 2.89 42 17.3 66 24.0 80 3.74 36 12.7 82 36.0 117 3.11 60 18.1 27 31.2 25 7.48 21 45.8 88 59.6 100 11.1 92 29.3 90 50.6 81 5.16 64 35.4 29 74.1 38 4.08 37 32.0 78 51.6 82 3.42 9
Fusion [6]60.6 11.6 30 24.3 41 2.89 42 15.6 38 21.9 38 3.83 42 11.0 10 23.7 46 3.37 65 21.0 77 33.4 51 7.62 50 44.1 51 56.3 53 10.1 29 30.3 109 54.1 110 5.45 101 38.0 72 83.7 88 4.08 37 34.0 105 54.7 104 3.56 65
RFlow [90]60.7 11.6 30 24.3 41 3.00 58 19.3 91 24.8 102 4.36 69 11.6 44 29.7 95 3.37 65 20.0 64 36.1 79 7.72 69 43.0 31 55.2 44 10.1 29 27.9 51 51.8 101 4.97 19 37.1 60 82.8 86 4.08 37 31.6 65 49.5 61 3.56 65
Sparse Occlusion [54]61.3 11.7 38 25.9 72 3.00 58 18.1 76 24.6 97 3.83 42 11.3 26 22.7 36 3.11 60 18.7 40 34.1 61 7.70 64 45.0 71 58.0 73 11.1 92 28.5 73 44.2 11 5.26 79 39.3 89 83.7 88 3.92 3 31.9 73 51.7 84 3.56 65
S2D-Matching [84]61.6 12.3 89 25.7 69 2.94 46 17.2 64 23.7 73 4.00 52 11.7 45 28.7 84 3.00 2 17.7 16 31.9 37 7.55 36 46.8 113 60.1 106 10.4 61 30.0 102 51.5 96 5.29 85 37.0 58 77.7 60 4.04 8 31.8 69 50.9 73 3.46 34
TC-Flow [46]62.2 12.0 60 30.3 107 2.89 42 16.8 58 23.4 66 3.92 51 11.7 45 21.4 24 3.00 2 19.5 53 36.1 79 8.12 93 46.5 104 59.8 103 11.3 114 27.0 31 48.4 51 5.26 79 35.5 32 74.6 42 4.04 8 33.3 97 54.5 103 3.51 49
HBM-GC [105]62.8 11.8 46 23.8 33 3.11 71 16.8 58 24.2 86 3.87 47 10.7 5 18.7 2 3.00 2 18.9 45 32.9 46 7.68 59 46.8 113 60.8 115 11.5 127 34.5 125 61.7 120 5.48 109 37.7 67 81.9 83 4.04 8 30.5 47 47.8 44 3.51 49
SVFilterOh [111]62.8 11.9 55 26.1 77 2.94 46 14.3 15 20.9 20 3.70 24 12.0 60 26.7 71 3.00 2 19.9 62 36.1 79 7.62 50 46.7 110 59.8 103 11.4 121 30.7 112 55.1 113 5.07 45 36.0 45 77.2 57 4.04 8 32.4 85 53.2 98 3.51 49
Black & Anandan [4]62.9 12.3 89 24.0 35 3.46 110 21.2 120 25.4 112 5.35 101 18.1 107 25.0 57 5.35 110 24.4 106 34.9 66 7.77 73 42.2 16 53.5 21 10.1 29 26.9 30 46.5 29 4.97 19 39.5 90 77.2 57 4.08 37 29.3 20 42.8 11 3.56 65
Classic+CPF [83]64.2 12.2 80 24.6 53 2.83 16 15.6 38 22.1 41 3.74 36 12.0 60 30.7 100 3.00 2 17.7 16 30.9 24 7.44 17 47.2 119 61.3 119 11.2 99 31.2 116 55.9 114 5.26 79 39.9 96 88.8 98 4.04 8 33.6 100 54.0 101 3.42 9
FESL [72]65.0 12.2 80 25.1 65 2.83 16 14.9 26 21.6 31 3.70 24 12.1 73 33.7 111 3.00 2 19.7 56 35.0 68 7.72 69 46.2 98 60.2 110 11.3 114 29.3 90 50.4 76 5.32 89 39.6 91 88.6 97 3.92 3 32.4 85 51.2 76 3.42 9
CostFilter [40]65.3 13.1 110 33.1 114 2.71 2 15.2 30 21.3 25 3.56 4 14.0 89 42.7 125 3.00 2 22.0 89 44.4 118 7.26 2 45.8 88 57.2 64 10.4 61 27.2 36 48.1 49 5.45 101 39.9 96 89.4 100 4.08 37 35.6 112 56.1 111 3.37 2
Efficient-NL [60]65.5 11.8 46 23.8 33 2.83 16 16.7 56 23.3 64 3.70 24 18.4 109 29.0 87 3.70 80 19.4 51 34.0 60 7.51 25 45.1 73 58.5 82 11.1 92 30.0 102 51.5 96 5.07 45 40.1 99 88.9 99 4.08 37 33.0 94 52.4 93 3.42 9
EpicFlow [102]65.7 11.9 55 27.6 87 2.83 16 16.0 42 22.2 42 3.79 38 11.8 58 21.7 28 3.00 2 21.3 82 42.9 113 7.85 78 46.3 101 59.4 96 11.2 99 27.4 40 47.5 40 5.16 64 38.2 74 76.6 51 4.12 80 35.2 111 58.0 114 3.56 65
Bartels [41]65.8 12.2 80 29.9 103 3.37 103 17.4 67 24.3 88 4.83 87 11.3 26 24.7 56 3.70 80 21.2 79 35.4 71 9.15 113 41.3 7 51.0 6 9.87 4 29.7 99 50.2 69 6.32 126 33.7 5 70.7 10 4.20 94 30.2 40 48.4 47 3.79 119
2D-CLG [1]65.9 11.6 30 24.1 37 3.11 71 19.4 94 23.3 64 6.24 114 18.7 110 24.3 52 4.69 104 22.4 93 31.8 35 8.66 106 43.3 42 56.1 52 10.4 61 26.0 14 44.2 11 5.35 96 40.2 100 91.5 110 4.20 94 29.6 28 44.5 15 3.51 49
Filter Flow [19]66.1 11.8 46 23.1 25 3.37 103 20.0 105 25.1 107 5.23 98 12.2 75 26.0 63 3.70 80 22.1 90 32.7 44 7.94 83 42.1 14 51.9 11 10.4 61 28.1 57 49.0 58 5.07 45 38.4 77 81.6 80 4.16 88 30.0 34 45.5 26 3.74 116
SRR-TVOF-NL [91]66.4 12.9 107 28.7 94 3.00 58 16.9 60 23.1 60 4.69 80 11.5 43 27.0 73 3.00 2 22.2 91 37.3 88 7.59 44 44.8 68 57.9 71 11.0 86 29.1 87 51.9 102 4.90 6 35.7 37 77.7 60 4.08 37 33.0 94 51.5 81 3.56 65
Steered-L1 [118]66.7 11.2 10 22.6 22 2.89 42 16.2 46 22.6 48 4.55 76 21.7 113 32.4 107 5.00 107 23.4 100 38.3 95 10.7 117 44.7 64 57.4 65 9.88 5 28.0 54 48.5 53 5.32 89 37.1 60 79.2 64 4.12 80 31.4 59 51.1 75 3.51 49
Occlusion-TV-L1 [63]69.5 11.6 30 25.0 64 3.11 71 19.8 102 26.0 120 4.83 87 11.3 26 23.0 40 3.46 77 22.5 96 43.0 115 7.94 83 43.0 31 54.8 39 9.88 5 28.0 54 50.7 84 5.32 89 39.6 91 76.6 51 4.62 117 31.5 63 50.5 69 3.56 65
EPPM w/o HM [88]70.2 12.7 104 30.9 109 2.71 2 16.1 43 23.1 60 3.70 24 17.7 105 42.4 124 3.70 80 21.3 82 42.5 112 7.70 64 43.0 31 53.1 18 10.3 52 30.2 107 57.1 116 4.97 19 38.5 80 89.6 101 4.12 80 32.4 85 51.3 78 3.42 9
OFH [38]70.4 12.0 60 27.3 83 3.00 58 18.1 76 23.4 66 4.20 63 12.4 80 32.7 108 3.00 2 18.6 38 35.4 71 7.35 7 46.5 104 60.1 106 10.8 79 27.5 43 47.2 34 5.26 79 41.1 103 81.1 77 4.20 94 35.7 113 56.1 111 3.46 34
CNN-flow-warp+ref [117]70.6 11.0 3 22.4 19 3.11 71 17.6 69 22.9 55 5.92 109 16.1 97 28.3 83 4.00 95 23.5 101 30.2 15 10.7 117 44.8 68 58.5 82 11.3 114 26.5 26 46.5 29 5.29 85 41.5 107 91.5 110 4.32 106 30.4 45 47.5 40 3.51 49
Adaptive [20]71.0 11.6 30 26.7 80 3.11 71 20.2 109 25.9 117 5.07 92 12.0 60 23.0 40 3.37 65 20.4 71 36.6 85 7.77 73 44.3 55 58.5 82 9.98 13 28.3 64 49.1 60 5.16 64 42.5 112 90.6 107 4.08 37 31.6 65 48.8 54 3.65 99
Horn & Schunck [3]72.5 12.1 73 23.7 30 3.32 99 21.4 122 25.6 115 5.89 108 17.0 101 28.2 82 5.35 110 27.3 113 37.9 91 8.04 88 42.4 19 54.3 28 10.3 52 26.2 17 44.7 15 5.07 45 40.9 102 81.7 81 4.20 94 30.2 40 44.3 14 3.70 102
IAOF [50]73.2 13.0 108 29.2 97 3.37 103 23.7 127 27.4 129 6.45 118 16.4 99 28.7 84 3.46 77 22.7 97 33.1 49 8.37 100 43.4 43 55.6 48 10.0 17 27.6 45 50.1 67 4.97 19 38.3 75 82.7 85 4.08 37 30.0 34 46.8 34 3.56 65
TV-L1-improved [17]74.0 11.5 24 25.4 66 3.11 71 20.1 108 26.0 120 5.26 99 16.8 100 19.7 8 4.04 100 19.5 53 32.3 39 7.79 76 43.8 48 56.5 54 10.0 17 28.9 81 51.1 93 5.07 45 43.2 115 98.9 121 4.43 112 31.4 59 50.2 64 3.70 102
TriFlow [95]74.1 12.5 100 36.7 118 3.00 58 18.7 81 24.5 94 4.76 82 11.7 45 28.1 81 3.00 2 21.7 87 41.4 109 7.62 50 46.8 113 60.4 112 11.2 99 29.9 100 51.7 100 4.97 19 37.8 70 76.9 53 4.08 37 31.7 68 48.6 51 3.51 49
HBpMotionGpu [43]75.0 12.3 89 32.0 112 3.79 118 20.6 118 25.4 112 6.00 111 11.3 26 26.1 65 3.00 2 23.2 99 44.0 117 7.85 78 44.3 55 56.9 58 10.8 79 29.0 82 53.5 109 5.26 79 34.9 25 69.8 6 4.04 8 31.8 69 51.4 80 3.70 102
Nguyen [33]75.3 12.0 60 25.9 72 3.37 103 21.2 120 24.5 94 6.27 115 12.7 82 28.0 79 3.70 80 23.8 102 34.7 64 8.58 104 43.0 31 54.7 36 10.1 29 27.7 50 50.7 84 4.97 19 43.4 117 93.7 113 4.43 112 30.2 40 47.4 37 3.56 65
BriefMatch [124]75.8 12.1 73 29.2 97 3.11 71 16.5 53 22.5 46 6.61 120 18.0 106 22.7 36 5.69 113 26.2 109 35.5 75 18.2 127 43.7 46 54.7 36 10.4 61 29.6 97 50.2 69 5.94 123 35.8 39 72.5 26 4.16 88 32.1 80 50.2 64 3.56 65
GraphCuts [14]76.0 13.9 117 30.2 106 3.32 99 16.4 49 22.5 46 4.36 69 33.4 125 24.1 51 5.35 110 22.3 92 34.7 64 7.87 81 44.5 61 57.0 59 9.98 13 28.3 64 50.3 74 4.90 6 38.5 80 88.2 94 4.20 94 33.9 104 53.6 100 3.56 65
FlowNet2 [122]77.2 19.1 125 47.5 126 3.11 71 17.1 61 24.1 83 4.55 76 14.2 91 29.8 97 3.37 65 23.8 102 42.9 113 8.33 97 45.9 92 58.1 75 10.6 72 27.6 45 49.4 62 4.93 12 39.2 87 81.0 76 4.08 37 31.6 65 49.2 57 3.56 65
TI-DOFE [24]77.6 12.7 104 27.6 87 3.87 122 22.2 125 25.3 109 6.66 121 14.1 90 25.3 59 4.36 102 27.7 115 38.7 99 9.06 111 42.7 25 53.6 24 10.1 29 26.8 28 48.8 57 4.97 19 38.3 75 76.0 45 4.24 104 31.9 73 44.7 18 3.87 121
ROF-ND [107]80.2 12.4 96 24.4 48 2.83 16 17.9 72 23.9 77 4.08 59 12.0 60 26.6 70 3.00 2 29.5 119 48.9 122 8.72 107 45.4 79 58.6 86 11.1 92 31.1 115 53.4 107 5.26 79 38.9 85 74.2 39 4.20 94 38.0 116 60.3 119 3.56 65
NL-TV-NCC [25]81.4 13.7 114 27.3 83 2.94 46 18.5 78 24.7 99 4.04 54 15.0 93 29.0 87 3.70 80 25.6 108 46.4 120 7.94 83 42.0 13 51.9 11 10.4 61 30.6 111 51.9 102 5.29 85 41.9 109 81.7 81 4.40 108 31.3 58 48.6 51 3.79 119
TriangleFlow [30]81.8 12.5 100 25.9 72 3.11 71 18.8 83 24.3 88 4.24 64 13.2 86 29.7 95 3.46 77 21.2 79 35.4 71 7.94 83 44.4 59 57.7 67 9.95 7 29.4 95 48.6 54 5.07 45 43.9 118 99.9 122 4.43 112 42.1 124 69.7 127 3.56 65
Complementary OF [21]82.1 12.4 96 34.5 117 2.83 16 16.4 49 23.5 70 3.79 38 30.7 119 32.2 106 7.05 121 19.9 62 43.9 116 7.44 17 46.9 116 60.4 112 10.7 76 28.1 57 47.7 44 5.23 75 41.1 103 80.3 72 4.12 80 42.0 123 62.0 121 3.56 65
Correlation Flow [75]82.4 12.6 103 28.0 90 2.71 2 20.0 105 25.8 116 4.36 69 11.3 26 22.3 33 3.00 2 20.7 75 38.6 98 7.72 69 45.7 84 59.0 92 10.3 52 33.4 121 60.4 119 5.45 101 45.6 122 99.9 122 4.40 108 33.4 98 54.9 106 3.56 65
LocallyOriented [52]82.9 12.2 80 28.1 91 3.27 97 20.5 116 25.9 117 5.07 92 14.3 92 30.0 98 3.37 65 24.2 105 41.7 110 7.66 57 44.7 64 57.1 63 10.1 29 28.8 77 47.4 38 5.48 109 42.4 110 80.6 74 4.12 80 32.4 85 51.2 76 3.56 65
IAOF2 [51]83.8 12.7 104 28.7 94 3.32 99 20.4 114 25.9 117 4.76 82 12.7 82 31.7 104 3.11 60 22.4 93 35.8 76 8.06 92 45.9 92 59.6 100 10.8 79 29.9 100 51.5 96 5.10 61 39.0 86 79.7 68 4.08 37 31.2 55 49.0 56 3.56 65
Aniso-Texture [82]84.8 11.5 24 24.1 37 2.83 16 20.2 109 26.0 120 4.97 90 20.0 112 24.4 54 3.37 65 26.9 111 50.7 123 9.11 112 46.1 96 60.5 114 11.4 121 32.7 120 62.2 124 5.94 123 37.3 63 80.2 70 4.04 8 34.1 106 55.0 107 3.42 9
ACK-Prior [27]86.5 12.5 100 29.7 101 2.83 16 16.1 43 22.7 50 4.00 52 25.6 115 27.7 75 5.72 115 22.4 93 36.0 78 10.7 117 45.7 84 59.3 94 11.4 121 31.8 118 50.6 81 5.35 96 38.8 84 79.9 69 4.16 88 33.5 99 51.7 84 3.70 102
Rannacher [23]87.4 11.7 38 28.7 94 3.16 96 20.4 114 26.3 124 5.07 92 19.0 111 26.0 63 4.80 106 19.8 58 38.1 93 7.79 76 44.5 61 57.4 65 10.1 29 29.0 82 50.3 74 5.20 69 42.6 113 97.0 118 4.40 108 33.7 102 55.9 110 3.70 102
Learning Flow [11]88.6 12.1 73 24.6 53 3.27 97 19.7 99 25.2 108 5.00 91 39.7 127 47.7 128 7.68 123 24.6 107 35.0 68 8.19 96 45.2 76 58.6 86 10.5 71 28.4 66 48.0 47 5.45 101 38.4 77 77.8 62 4.40 108 32.6 92 48.4 47 3.92 123
2bit-BM-tele [98]89.5 11.7 38 27.0 82 3.79 118 20.2 109 26.3 124 5.07 92 12.0 60 23.2 44 4.00 95 21.2 79 36.1 79 8.16 95 45.3 77 58.0 73 10.3 52 34.0 124 61.8 121 5.92 122 54.1 129 99.9 122 5.72 127 29.8 32 47.4 37 3.74 116
FOLKI [16]90.1 13.0 108 30.9 109 4.97 127 22.2 125 24.9 105 9.00 127 17.3 103 33.0 110 7.00 118 33.4 124 38.7 99 17.0 125 44.3 55 55.8 49 10.4 61 27.6 45 49.7 63 5.48 109 36.2 51 74.2 39 4.80 120 30.4 45 44.9 21 4.08 125
SimpleFlow [49]90.7 12.0 60 24.0 35 2.94 46 18.5 78 24.4 91 4.24 64 32.7 122 39.0 119 5.69 113 18.0 25 36.2 83 7.55 36 46.9 116 60.8 115 11.1 92 31.4 117 58.1 117 5.35 96 49.4 126 99.9 122 5.16 125 40.0 120 63.0 124 3.46 34
SILK [79]91.1 13.3 112 30.7 108 3.83 121 22.0 124 25.3 109 7.16 122 34.7 126 40.0 121 7.77 125 26.6 110 36.6 85 8.60 105 45.1 73 57.9 71 10.0 17 28.4 66 50.9 90 6.03 125 34.8 19 71.8 21 4.51 116 31.4 59 48.0 45 3.74 116
StereoFlow [44]92.5 22.8 129 48.3 127 3.74 117 20.5 116 26.8 127 5.07 92 11.3 26 29.3 92 3.37 65 20.1 66 37.0 87 7.62 50 59.3 127 75.2 127 10.8 79 39.3 129 71.4 128 5.45 101 35.8 39 73.9 35 4.08 37 35.7 113 55.1 109 3.70 102
StereoOF-V1MT [119]94.0 13.7 114 32.7 113 3.00 58 18.7 81 23.6 71 4.80 85 21.8 114 28.0 79 5.07 109 31.6 120 40.6 106 9.57 114 46.5 104 58.9 91 11.5 127 29.2 89 50.2 69 6.45 128 42.4 110 94.7 114 4.80 120 31.4 59 48.3 46 3.46 34
Shiralkar [42]94.6 13.2 111 31.6 111 3.00 58 19.7 99 24.5 94 4.65 79 17.0 101 30.7 100 4.08 101 32.1 122 53.1 125 8.04 88 46.3 101 59.7 102 10.3 52 28.4 66 50.2 69 5.45 101 45.5 121 95.2 116 4.24 104 39.2 119 62.6 122 3.42 9
Dynamic MRF [7]96.5 12.1 73 26.8 81 2.94 46 18.0 75 23.9 77 4.16 62 18.3 108 30.7 100 5.00 107 28.9 116 39.8 102 10.5 116 45.9 92 58.6 86 11.2 99 30.9 113 56.0 115 5.80 121 43.0 114 90.3 105 4.65 118 33.7 102 51.8 86 3.70 102
Adaptive flow [45]97.2 13.4 113 25.8 71 4.51 124 21.8 123 25.4 112 7.26 123 13.7 88 27.5 74 4.69 104 24.1 104 35.2 70 8.76 109 47.3 120 61.5 120 10.2 47 33.8 123 61.9 122 5.45 101 35.9 43 73.2 32 4.20 94 34.7 110 54.7 104 3.70 102
UnFlow [129]98.5 14.9 120 40.2 123 3.11 71 18.5 78 23.4 66 5.48 103 15.3 94 31.3 103 4.36 102 22.8 98 38.0 92 8.45 102 48.3 123 63.0 123 10.9 85 32.4 119 62.0 123 5.72 116 35.4 29 71.2 15 4.32 106 45.5 127 66.0 126 3.87 121
SPSA-learn [13]101.6 12.3 89 33.7 116 3.37 103 19.2 89 23.6 71 5.45 102 30.0 118 39.7 120 7.00 118 26.9 111 41.3 108 8.41 101 46.7 110 60.1 106 10.2 47 29.4 95 50.6 81 5.20 69 53.7 128 99.9 122 8.43 129 51.4 128 72.0 128 3.51 49
HCIC-L [99]103.3 21.0 128 41.8 124 5.07 128 20.2 109 26.1 123 5.80 106 16.3 98 42.3 123 4.00 95 31.7 121 51.0 124 8.50 103 44.7 64 55.5 46 10.4 61 35.2 126 69.8 127 5.07 45 39.9 96 91.2 109 4.16 88 40.4 122 58.0 114 3.65 99
SegOF [10]103.5 12.3 89 33.1 114 3.11 71 17.9 72 23.8 75 4.51 74 29.0 117 34.3 113 6.16 116 32.8 123 78.9 129 8.33 97 48.1 122 63.6 124 11.2 99 28.5 73 54.3 111 5.72 116 44.6 119 99.9 122 4.97 123 37.9 115 61.4 120 3.51 49
FFV1MT [106]104.9 17.0 123 37.6 121 3.37 103 19.3 91 22.9 55 6.40 117 28.2 116 46.7 127 6.95 117 29.3 117 38.4 96 11.4 121 46.3 101 58.2 79 10.4 61 29.0 82 50.4 76 5.72 116 46.7 123 88.5 96 4.93 122 39.0 118 56.9 113 4.43 127
PGAM+LK [55]105.3 15.5 122 39.4 122 4.55 125 19.8 102 24.0 80 7.68 124 33.1 124 43.4 126 8.00 126 34.5 126 45.7 119 11.2 120 46.6 108 57.7 67 10.6 72 29.3 90 50.8 89 5.74 120 37.4 64 77.2 57 4.43 112 34.4 108 53.3 99 4.24 126
Heeger++ [104]105.6 19.8 126 44.7 125 3.11 71 18.9 85 22.8 52 6.45 118 33.0 123 35.2 114 7.16 122 29.3 117 38.4 96 11.4 121 51.5 125 65.2 125 11.3 114 28.4 66 46.9 31 6.78 129 47.9 125 84.5 90 4.69 119 40.1 121 58.8 117 3.70 102
SLK [47]107.2 13.9 117 29.9 103 3.79 118 20.0 105 22.8 52 6.22 113 32.0 121 33.7 111 7.72 124 33.4 124 46.4 120 16.1 124 48.5 124 61.7 121 10.3 52 28.4 66 47.9 46 5.72 116 43.2 115 97.9 119 4.97 123 38.7 117 59.8 118 4.04 124
Pyramid LK [2]115.0 14.4 119 37.3 120 4.93 126 23.7 127 25.3 109 9.98 129 42.2 128 35.7 115 12.3 128 56.2 129 64.2 127 35.8 129 65.6 128 83.9 128 10.6 72 28.1 57 46.1 25 5.48 109 45.2 120 99.9 122 5.89 128 53.6 129 75.1 129 5.42 128
GroupFlow [9]115.5 19.9 127 49.6 128 3.42 109 19.1 87 24.1 83 5.48 103 31.4 120 40.0 121 8.19 127 36.2 127 61.9 126 12.1 123 55.6 126 71.3 126 11.4 121 36.3 127 67.0 126 5.60 115 46.7 123 98.5 120 4.20 94 43.6 125 62.8 123 3.56 65
Periodicity [78]127.7 17.6 124 55.7 129 5.45 129 26.8 129 27.0 128 9.75 128 49.4 129 51.5 129 17.7 129 51.3 128 70.3 128 27.9 128 66.6 129 86.3 129 11.7 129 38.7 128 82.5 129 6.38 127 51.8 127 99.9 122 5.48 126 44.3 126 65.5 125 5.80 129
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 T. Arici. Energy minimization based motion estimation using adaptive smoothness priors. Submitted to IEEE TIP 2011.
[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 D. Nguyen. Enhancing the sharpness of flow field using image-driven functions with occlusion-aware filter. Submitted to TIP 2011.
[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 A. Ayvaci, M. Raptis, and S. Soatto. Sparse occlusion detection with optical flow. Submitted to IJCV 2011.
[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 L. Chen, J. Wang, and Y. Wu. Decomposing and regularizing sparse/non-sparse components for motion field estimation. Submitted to PAMI 2013.
[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 M. Santoro, G. AlRegib, and Y. Altunbasak. Motion estimation using block overlap minimization. Submitted to 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 W. Dong, G. Shi, X. Hu, and Y. Ma. Nonlocal sparse and low-rank regularization for optical flow estimation. Submitted to IEEE TIP 2013.
[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] Correlation Flow 290 2 color M. Drulea and S. Nedevschi. Motion estimation using the correlation transform. TIP 2013. Matlab code.
[76] TC/T-Flow 341 5 color M. Stoll, S. Volz, and A. Bruhn. Joint trilateral filtering for multiframe optical flow. ICIP 2013.
[77] OFLAF 1530 2 color T. Kim, H. Lee, and K. Lee. Optical flow via locally adaptive fusion of complementary data costs. ICCV 2013.
[78] Periodicity 8000 4 color G. Khachaturov, S. Gonzalez-Brambila, and J. Gonzalez-Trejo. Periodicity-based computation of optical flow. Submitted to Computacion y Sistemas (CyS) 2013.
[79] SILK 572 2 gray P. Zille, C. Xu, T. Corpetti, L. Shao. Observation models based on scale interactions for optical flow estimation. Submitted to IEEE TIP.
[80] 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.
[81] SuperFlow 178 2 color Anonymous. Superpixel based optical flow estimation. ICCV 2013 submission 507.
[82] Aniso-Texture 300 2 color Anonymous. Texture information-based optical flow estimation using an incremental multi-resolution approach. ITC-CSCC 2013 submission 267.
[83] Classic+CPF 640 2 gray Z. Tu, R. Veltkamp, and N. van der Aa. A combined post-filtering method to improve accuracy of variational optical flow estimation. Submitted to Pattern Recognition 2013.
[84] S2D-Matching 1200 2 color Anonymous. Locally affine sparse-to-dense matching for motion and occlusion estimation. ICCV 2013 submission 1479.
[85] AGIF+OF 438 2 gray Z. Tu, R. Poppe, and R. Veltkamp. Adaptive guided image filter to warped interpolation image for variational optical flow computation. Submitted to Signal Processing 2015.
[86] DeepFlow 13 2 color P. Weinzaepfel, J. Revaud, Z. Harchaoui, and C. Schmid. DeepFlow: large displacement optical flow with deep matching. ICCV 2013.
[87] NNF-Local 673 2 color Z. Chen, H. Jin, Z. Lin, S. Cohen, and Y. Wu. Large displacement optical flow with nearest neighbor field. Submitted to PAMI 2014.
[88] 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.
[89] 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.
[90] 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.
[91] 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.
[92] 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.
[93] 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.
[94] FMOF 215 2 color N. Jith, A. Ramakanth, and V. Babu. Optical flow estimation using approximate nearest neighbor field fusion. ICASSP 2014.
[95] TriFlow 150 2 color TriFlow. Optical flow with geometric occlusion estimation and fusion of multiple frames. ECCV 2014 submission 914.
[96] ComponentFusion 6.5 2 color Anonymous. Fast optical flow by component fusion. ECCV 2014 submission 941.
[97] 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.
[98] 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.
[99] HCIC-L 330 2 color Anonymous. Globally-optimal image correspondence using a hierarchical graphical model. NIPS 2014 submission 114.
[100] TF+OM 600 2 color R. Kennedy and C. Taylor. Optical flow with geometric occlusion estimation and fusion of multiple frames. EMMCVPR 2015.
[101] PH-Flow 800 2 color J. Yang and H. Li. Dense, accurate optical flow estimation with piecewise parametric model. CVPR 2015.
[102] EpicFlow 16 2 color J. Revaud, P. Weinzaepfel, Z. Harchaoui, and C. Schmid. EpicFlow: edge-preserving interpolation of correspondences for optical flow. CVPR 2015.
[103] NNF-EAC 380 2 color Anonymous. Variational method for joint optical flow estimation and edge-aware image restoration. CVPR 2015 submission 2336.
[104] Heeger++ 6600 5 gray Anonymous. A context aware biologically inspired algorithm for optical flow (updated results). CVPR 2015 submission 2238.
[105] HBM-GC 330 2 color A. Zheng and Y. Yuan. Motion estimation via hierarchical block matching and graph cut. Submitted to ICIP 2015.
[106] 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.
[107] 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.
[108] DeepFlow2 16 2 color J. Revaud, P. Weinzaepfel, Z. Harchaoui, and C. Schmid. Deep convolutional matching. Submitted to IJCV, 2015.
[109] HAST 2667 2 color Anonymous. Highly accurate optical flow estimation on superpixel tree. ICCV 2015 submission 2221.
[110] FlowFields 15 2 color Anonymous. Flow Fields: Dense unregularized correspondence fields for highly accurate large displacement optical flow estimation. ICCV 2015 submission 744.
[111] SVFilterOh 1.56 2 color Anonymous. Fast estimation of large displacement optical flow using PatchMatch and dominant motion patterns. CVPR 2016 submission 1788.
[112] FlowNetS+ft+v 0.5 2 color Anonymous. Learning optical flow with convolutional neural networks. ICCV 2015 submission 235.
[113] 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.)
[114] 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.
[115] DF-Auto 70 2 color N. Monzon, A. Salgado, and J. Sanchez. Regularization strategies for discontinuity-preserving optical flow methods. Submitted to TIP 2015.
[116] CPM-Flow 3 2 color Anonymous. Efficient coarse-to-fine PatchMatch for large displacement optical flow. CVPR 2016 submission 241.
[117] 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.
[118] Steered-L1 804 2 color Anonymous. Optical flow estimation via steered-L1 norm. Submitted to WSCG 2016.
[119] 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.
[120] PGM-C 5 2 color Y. Li. Pyramidal gradient matching for optical flow estimation. Submitted to PAMI 2016.
[121] 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. Submitted to TIP 2016.
[122] FlowNet2 0.091 2 color Anonymous. FlowNet 2.0: Evolution of optical flow estimation with deep networks. CVPR 2017 submission 900.
[123] S2F-IF 20 2 color Anonymous. S2F-IF: Slow-to-fast interpolator flow. CVPR 2017 submission 765.
[124] 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.
[125] OAR-Flow 60 2 color Anonymous. Order-adaptive regularisation for variational optical flow: global, local and in between. SSVM 2017 submission 20.
[126] AdaConv-v1 2.8 2 color S. Niklaus, L. Mai, and F. Liu. (Interpolation results only.) Video frame interpolation via adaptive convolution. CVPR 2017.
[127] SepConv-v1 0.2 2 color S. Niklaus, L. Mai, and F. Liu. (Interpolation results only.) Video frame interpolation via adaptive separable convolution. ICCV 2017.
[128] ProbFlowFields 37 2 color A. Wannenwetsch, M. Keuper, and S. Roth. ProbFlow: joint optical flow and uncertainty estimation. ICCV 2017.
[129] UnFlow 0.12 2 color Anonymous. UnFlow: Unsupervised learning of optical flow with a bidirectional census loss. Submitted to AAAI 2018.
* 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.