Optical flow evaluation results |
Statistics:
Average
SD
R0.5
R1.0
R2.0
A90
A95
A99
Error type: endpoint angle interpolation normalized interpolation |
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 | |
MDP-Flow2 [68] | 9.9 | 35.7 2 | 30.6 2 | 47.8 8 | 25.9 10 | 30.5 10 | 36.9 3 | 28.6 3 | 29.8 5 | 38.5 2 | 51.9 7 | 46.5 15 | 80.3 10 | 71.9 4 | 66.6 3 | 87.2 8 | 68.6 5 | 53.9 23 | 82.1 35 | 28.1 2 | 43.6 8 | 42.4 10 | 36.6 28 | 55.6 29 | 50.0 6 |
PMMST [114] | 10.9 | 35.8 4 | 30.8 3 | 47.9 18 | 26.5 32 | 31.0 20 | 37.3 36 | 28.6 3 | 29.9 6 | 38.4 1 | 51.7 3 | 46.0 5 | 80.2 5 | 72.0 9 | 66.7 8 | 87.3 16 | 68.5 1 | 53.3 7 | 82.0 13 | 28.1 2 | 43.7 13 | 42.4 10 | 36.5 18 | 55.5 23 | 50.0 6 |
PH-Flow [101] | 11.7 | 36.1 21 | 32.5 29 | 47.8 8 | 25.6 4 | 29.6 4 | 36.9 3 | 28.7 9 | 30.0 8 | 38.5 2 | 51.6 1 | 45.5 1 | 80.2 5 | 71.9 4 | 66.7 8 | 87.3 16 | 68.8 35 | 54.8 66 | 81.9 5 | 28.1 2 | 43.6 8 | 42.4 10 | 36.4 12 | 55.3 14 | 50.0 6 |
NNF-Local [87] | 12.1 | 35.7 2 | 31.4 6 | 47.6 3 | 25.5 2 | 29.6 4 | 36.9 3 | 28.6 3 | 29.9 6 | 38.5 2 | 52.4 31 | 48.0 62 | 80.3 10 | 72.0 9 | 66.6 3 | 87.4 34 | 68.7 19 | 54.4 45 | 82.0 13 | 28.1 2 | 43.5 6 | 42.4 10 | 36.2 4 | 55.0 6 | 50.0 6 |
NN-field [71] | 17.1 | 36.0 16 | 32.2 18 | 47.9 18 | 25.5 2 | 29.3 3 | 36.8 1 | 29.4 60 | 29.7 4 | 39.0 43 | 52.4 31 | 48.1 68 | 80.3 10 | 72.0 9 | 66.7 8 | 87.3 16 | 68.7 19 | 54.0 28 | 82.0 13 | 28.1 2 | 43.4 4 | 42.4 10 | 36.4 12 | 55.2 10 | 50.0 6 |
IROF++ [58] | 20.9 | 36.2 28 | 33.0 43 | 47.8 8 | 26.1 15 | 30.9 17 | 36.9 3 | 29.1 30 | 31.0 29 | 38.9 28 | 51.6 1 | 45.6 2 | 80.4 19 | 72.0 9 | 66.8 17 | 87.2 8 | 68.6 5 | 53.4 8 | 82.2 57 | 28.3 17 | 44.6 39 | 42.4 10 | 36.5 18 | 55.3 14 | 50.4 76 |
ProbFlowFields [128] | 21.1 | 35.9 9 | 32.4 23 | 48.0 29 | 25.8 8 | 30.5 10 | 37.2 31 | 28.6 3 | 30.3 11 | 38.5 2 | 52.1 20 | 46.4 11 | 80.7 56 | 72.3 64 | 67.1 65 | 87.5 78 | 68.6 5 | 53.8 15 | 82.1 35 | 28.0 1 | 42.8 1 | 42.3 1 | 36.1 2 | 54.6 2 | 50.1 24 |
Sparse-NonSparse [56] | 22.5 | 36.2 28 | 32.8 37 | 48.0 29 | 25.9 10 | 30.4 8 | 37.0 11 | 29.0 22 | 30.9 26 | 38.8 13 | 52.0 13 | 46.1 7 | 80.6 41 | 72.1 24 | 66.8 17 | 87.3 16 | 68.9 45 | 54.6 56 | 82.1 35 | 28.3 17 | 44.0 21 | 42.4 10 | 36.4 12 | 55.4 18 | 50.1 24 |
CombBMOF [113] | 23.8 | 35.9 9 | 31.0 4 | 47.8 8 | 25.8 8 | 30.5 10 | 36.8 1 | 29.2 37 | 30.8 23 | 39.5 80 | 52.4 31 | 47.4 41 | 80.3 10 | 72.1 24 | 66.8 17 | 87.4 34 | 68.9 45 | 54.5 49 | 82.1 35 | 28.5 57 | 44.6 39 | 42.3 1 | 36.0 1 | 54.6 2 | 50.0 6 |
nLayers [57] | 25.7 | 36.4 53 | 32.0 13 | 48.2 57 | 26.0 13 | 30.4 8 | 37.3 36 | 28.7 9 | 29.4 2 | 38.8 13 | 52.2 27 | 46.8 22 | 80.4 19 | 72.3 64 | 67.1 65 | 87.4 34 | 68.8 35 | 54.7 60 | 82.0 13 | 28.3 17 | 43.7 13 | 42.4 10 | 36.4 12 | 55.4 18 | 49.9 3 |
AGIF+OF [85] | 25.8 | 36.2 28 | 32.8 37 | 47.9 18 | 26.1 15 | 30.8 14 | 37.1 18 | 29.0 22 | 30.7 19 | 38.9 28 | 51.8 5 | 46.2 9 | 80.1 3 | 72.3 64 | 67.2 79 | 87.3 16 | 68.9 45 | 55.2 82 | 81.9 5 | 28.3 17 | 43.6 8 | 42.4 10 | 36.6 28 | 56.0 46 | 49.9 3 |
2DHMM-SAS [92] | 27.5 | 36.4 53 | 33.9 80 | 47.9 18 | 27.1 57 | 32.6 43 | 37.0 11 | 28.5 2 | 30.4 14 | 38.9 28 | 51.8 5 | 45.6 2 | 80.4 19 | 72.1 24 | 66.9 29 | 87.4 34 | 68.8 35 | 54.5 49 | 82.0 13 | 28.3 17 | 44.2 27 | 42.3 1 | 36.7 40 | 56.1 54 | 50.0 6 |
NNF-EAC [103] | 28.1 | 36.3 42 | 32.4 23 | 48.0 29 | 26.6 35 | 31.7 30 | 37.1 18 | 29.3 47 | 30.2 9 | 39.0 43 | 52.4 31 | 46.9 25 | 81.1 97 | 72.0 9 | 66.7 8 | 87.4 34 | 68.7 19 | 53.7 14 | 82.1 35 | 28.2 11 | 43.9 16 | 42.4 10 | 36.7 40 | 55.9 43 | 50.0 6 |
Layers++ [37] | 28.7 | 36.3 42 | 32.4 23 | 48.2 57 | 25.7 5 | 29.2 2 | 37.3 36 | 28.9 18 | 30.6 16 | 38.9 28 | 52.0 13 | 46.4 11 | 80.4 19 | 72.2 36 | 67.0 42 | 87.5 78 | 68.9 45 | 55.2 82 | 82.0 13 | 28.3 17 | 44.0 21 | 42.4 10 | 36.6 28 | 55.5 23 | 50.1 24 |
LSM [39] | 30.1 | 36.3 42 | 33.7 70 | 48.0 29 | 26.1 15 | 31.0 20 | 37.0 11 | 29.1 30 | 31.8 44 | 38.9 28 | 52.2 27 | 46.9 25 | 80.6 41 | 72.1 24 | 66.9 29 | 87.3 16 | 69.0 59 | 54.9 69 | 82.1 35 | 28.3 17 | 44.1 25 | 42.4 10 | 36.5 18 | 55.7 33 | 50.0 6 |
ComponentFusion [96] | 30.5 | 36.0 16 | 32.2 18 | 48.0 29 | 26.1 15 | 31.1 25 | 36.9 3 | 29.1 30 | 32.3 52 | 38.8 13 | 52.0 13 | 47.0 29 | 80.3 10 | 72.2 36 | 67.1 65 | 87.3 16 | 68.7 19 | 53.9 23 | 82.1 35 | 28.5 57 | 46.1 92 | 42.4 10 | 36.7 40 | 55.8 40 | 50.2 45 |
FlowFields [110] | 30.5 | 36.0 16 | 32.7 34 | 47.9 18 | 26.4 29 | 32.0 32 | 37.3 36 | 29.0 22 | 32.6 58 | 38.7 8 | 52.5 36 | 47.9 59 | 80.7 56 | 72.3 64 | 67.0 42 | 87.5 78 | 68.6 5 | 54.4 45 | 82.0 13 | 28.2 11 | 44.0 21 | 42.4 10 | 36.3 6 | 55.2 10 | 50.1 24 |
S2F-IF [123] | 31.4 | 35.9 9 | 32.5 29 | 47.8 8 | 26.2 22 | 31.6 29 | 37.2 31 | 29.0 22 | 31.9 49 | 38.6 7 | 52.3 29 | 47.6 46 | 80.4 19 | 72.4 91 | 67.2 79 | 87.5 78 | 68.7 19 | 54.5 49 | 81.9 5 | 28.4 38 | 44.7 44 | 42.4 10 | 36.3 6 | 55.2 10 | 50.1 24 |
TV-L1-MCT [64] | 31.5 | 36.8 84 | 34.7 100 | 48.2 57 | 26.7 36 | 32.4 40 | 37.3 36 | 28.6 3 | 30.9 26 | 39.0 43 | 51.9 7 | 45.7 4 | 80.5 35 | 72.2 36 | 67.0 42 | 87.3 16 | 68.6 5 | 53.0 4 | 82.3 71 | 28.3 17 | 44.4 32 | 42.4 10 | 36.1 2 | 54.9 5 | 50.2 45 |
FlowFields+ [130] | 31.5 | 35.9 9 | 32.6 33 | 47.9 18 | 26.4 29 | 32.2 35 | 37.4 47 | 29.0 22 | 32.6 58 | 38.7 8 | 52.3 29 | 47.7 51 | 80.6 41 | 72.3 64 | 67.1 65 | 87.5 78 | 68.7 19 | 54.6 56 | 82.0 13 | 28.2 11 | 44.0 21 | 42.4 10 | 36.3 6 | 55.2 10 | 50.1 24 |
WLIF-Flow [93] | 32.3 | 36.1 21 | 32.5 29 | 47.8 8 | 26.3 27 | 31.2 26 | 37.1 18 | 29.1 30 | 30.7 19 | 39.1 50 | 52.0 13 | 46.4 11 | 80.6 41 | 72.1 24 | 66.8 17 | 87.4 34 | 69.0 59 | 54.9 69 | 82.3 71 | 28.3 17 | 43.9 16 | 42.5 62 | 36.8 47 | 55.9 43 | 50.1 24 |
LME [70] | 34.6 | 35.8 4 | 31.0 4 | 47.8 8 | 26.9 46 | 32.2 35 | 38.4 85 | 29.2 37 | 32.6 58 | 38.8 13 | 51.9 7 | 46.7 20 | 80.4 19 | 72.6 111 | 67.4 103 | 87.7 119 | 68.8 35 | 54.9 69 | 82.0 13 | 28.1 2 | 43.5 6 | 42.4 10 | 36.3 6 | 55.3 14 | 50.0 6 |
COFM [59] | 34.8 | 36.1 21 | 32.0 13 | 48.1 43 | 26.1 15 | 30.8 14 | 37.1 18 | 28.8 12 | 30.3 11 | 38.8 13 | 51.7 3 | 46.0 5 | 80.0 1 | 72.2 36 | 67.2 79 | 87.2 8 | 68.9 45 | 56.1 108 | 81.7 1 | 28.1 2 | 42.8 1 | 43.1 116 | 37.1 78 | 56.9 87 | 50.7 106 |
FMOF [94] | 35.6 | 36.5 61 | 33.7 70 | 48.2 57 | 25.9 10 | 30.3 7 | 37.1 18 | 29.3 47 | 30.7 19 | 39.0 43 | 52.5 36 | 47.5 42 | 80.2 5 | 72.2 36 | 67.0 42 | 87.5 78 | 69.0 59 | 55.1 79 | 82.0 13 | 28.1 2 | 43.4 4 | 42.4 10 | 36.8 47 | 56.0 46 | 50.1 24 |
DeepFlow2 [108] | 35.7 | 36.2 28 | 32.4 23 | 48.2 57 | 27.1 57 | 32.9 47 | 37.8 70 | 29.2 37 | 32.9 66 | 39.0 43 | 52.5 36 | 47.5 42 | 80.5 35 | 72.2 36 | 66.9 29 | 87.5 78 | 68.5 1 | 52.9 3 | 82.1 35 | 28.3 17 | 44.4 32 | 42.4 10 | 36.4 12 | 55.4 18 | 50.2 45 |
OFLAF [77] | 36.0 | 35.8 4 | 31.5 7 | 47.8 8 | 25.7 5 | 29.8 6 | 37.0 11 | 29.0 22 | 31.2 32 | 38.7 8 | 52.0 13 | 46.8 22 | 80.1 3 | 72.4 91 | 67.3 93 | 87.4 34 | 68.9 45 | 55.3 87 | 82.0 13 | 28.6 75 | 45.4 77 | 42.4 10 | 37.1 78 | 57.1 97 | 50.1 24 |
RNLOD-Flow [121] | 36.1 | 36.3 42 | 33.5 59 | 48.0 29 | 26.8 41 | 32.6 43 | 37.1 18 | 29.2 37 | 31.8 44 | 38.8 13 | 52.1 20 | 46.9 25 | 80.2 5 | 72.2 36 | 67.0 42 | 87.3 16 | 69.0 59 | 55.2 82 | 82.1 35 | 28.3 17 | 44.2 27 | 42.4 10 | 37.1 78 | 56.9 87 | 49.8 1 |
DeepFlow [86] | 37.2 | 36.1 21 | 31.8 11 | 48.1 43 | 27.3 62 | 32.9 47 | 38.4 85 | 29.3 47 | 33.3 80 | 39.1 50 | 52.6 49 | 47.0 29 | 80.7 56 | 72.2 36 | 66.8 17 | 87.5 78 | 68.7 19 | 52.8 2 | 82.5 96 | 28.1 2 | 43.6 8 | 42.3 1 | 36.2 4 | 55.0 6 | 50.2 45 |
Ramp [62] | 38.0 | 36.5 61 | 34.0 85 | 48.2 57 | 26.0 13 | 30.8 14 | 37.1 18 | 28.9 18 | 30.8 23 | 38.8 13 | 51.9 7 | 46.1 7 | 80.4 19 | 72.2 36 | 67.0 42 | 87.4 34 | 69.1 69 | 55.4 93 | 82.2 57 | 28.4 38 | 44.7 44 | 42.4 10 | 36.8 47 | 56.2 62 | 50.2 45 |
MDP-Flow [26] | 38.2 | 35.8 4 | 31.5 7 | 48.0 29 | 26.2 22 | 31.4 28 | 37.4 47 | 29.0 22 | 31.1 30 | 38.9 28 | 52.7 58 | 47.8 56 | 80.7 56 | 72.2 36 | 66.9 29 | 87.5 78 | 68.9 45 | 55.2 82 | 82.1 35 | 28.5 57 | 45.3 76 | 42.5 62 | 36.3 6 | 55.4 18 | 50.0 6 |
IROF-TV [53] | 39.3 | 36.3 42 | 33.6 66 | 48.2 57 | 26.2 22 | 31.0 20 | 37.0 11 | 29.3 47 | 33.6 86 | 39.1 50 | 51.9 7 | 46.5 15 | 80.8 69 | 72.3 64 | 67.0 42 | 87.6 110 | 68.5 1 | 53.9 23 | 81.9 5 | 28.3 17 | 44.9 56 | 42.3 1 | 36.6 28 | 55.6 29 | 50.4 76 |
PGM-C [120] | 39.5 | 36.2 28 | 33.3 52 | 48.1 43 | 26.5 32 | 32.2 35 | 37.5 56 | 29.2 37 | 32.9 66 | 38.8 13 | 52.5 36 | 48.3 82 | 80.7 56 | 72.3 64 | 67.0 42 | 87.5 78 | 68.6 5 | 54.0 28 | 82.0 13 | 28.3 17 | 44.6 39 | 42.4 10 | 36.5 18 | 55.5 23 | 50.4 76 |
Classic+NL [31] | 39.7 | 36.5 61 | 34.0 85 | 48.2 57 | 26.2 22 | 30.9 17 | 37.1 18 | 28.8 12 | 30.6 16 | 38.8 13 | 52.1 20 | 46.5 15 | 80.6 41 | 72.2 36 | 67.0 42 | 87.4 34 | 69.2 79 | 55.3 87 | 82.2 57 | 28.4 38 | 44.6 39 | 42.4 10 | 36.8 47 | 56.2 62 | 50.2 45 |
DF-Auto [115] | 40.2 | 36.8 84 | 31.9 12 | 48.9 94 | 28.5 87 | 33.7 66 | 40.8 99 | 28.8 12 | 30.3 11 | 38.7 8 | 52.5 36 | 47.3 35 | 80.4 19 | 72.1 24 | 66.7 8 | 87.4 34 | 68.6 5 | 53.8 15 | 82.0 13 | 28.4 38 | 44.7 44 | 42.5 62 | 36.8 47 | 56.3 67 | 50.2 45 |
FC-2Layers-FF [74] | 42.5 | 36.4 53 | 33.8 76 | 48.1 43 | 25.7 5 | 29.1 1 | 37.4 47 | 28.9 18 | 30.9 26 | 38.8 13 | 52.1 20 | 46.8 22 | 80.6 41 | 72.3 64 | 67.2 79 | 87.4 34 | 69.1 69 | 55.5 95 | 82.1 35 | 28.4 38 | 44.7 44 | 42.5 62 | 36.9 61 | 56.3 67 | 50.0 6 |
SuperFlow [81] | 43.0 | 36.5 61 | 32.2 18 | 48.8 91 | 28.3 82 | 33.4 61 | 40.9 100 | 29.5 70 | 32.6 58 | 39.4 73 | 52.5 36 | 46.7 20 | 80.9 77 | 72.2 36 | 66.9 29 | 87.5 78 | 68.5 1 | 53.4 8 | 82.0 13 | 28.3 17 | 44.7 44 | 42.4 10 | 36.3 6 | 55.4 18 | 50.1 24 |
HAST [109] | 43.6 | 36.1 21 | 31.7 9 | 48.1 43 | 26.1 15 | 31.0 20 | 37.0 11 | 29.3 47 | 31.7 41 | 39.2 62 | 51.9 7 | 46.5 15 | 80.3 10 | 72.3 64 | 67.3 93 | 87.2 8 | 69.3 90 | 56.4 116 | 82.0 13 | 28.4 38 | 45.0 63 | 42.5 62 | 37.3 91 | 57.3 101 | 50.0 6 |
CPM-Flow [116] | 44.4 | 36.2 28 | 33.5 59 | 48.1 43 | 26.5 32 | 32.2 35 | 37.5 56 | 29.3 47 | 32.6 58 | 38.9 28 | 52.7 58 | 48.7 95 | 80.7 56 | 72.3 64 | 67.0 42 | 87.5 78 | 68.7 19 | 53.8 15 | 82.2 57 | 28.4 38 | 44.7 44 | 42.4 10 | 36.5 18 | 55.5 23 | 50.3 62 |
Second-order prior [8] | 44.5 | 36.2 28 | 32.1 16 | 48.1 43 | 27.9 76 | 34.1 72 | 37.4 47 | 29.9 86 | 34.6 101 | 39.7 90 | 52.4 31 | 47.2 33 | 80.6 41 | 71.9 4 | 66.6 3 | 87.5 78 | 68.7 19 | 54.0 28 | 82.1 35 | 28.5 57 | 45.2 74 | 42.4 10 | 36.5 18 | 55.7 33 | 50.2 45 |
Kuang [131] | 45.3 | 36.2 28 | 33.5 59 | 47.9 18 | 27.0 52 | 33.3 58 | 37.4 47 | 29.4 60 | 32.9 66 | 39.1 50 | 52.5 36 | 48.1 68 | 80.6 41 | 72.3 64 | 67.1 65 | 87.4 34 | 68.7 19 | 54.3 40 | 82.1 35 | 28.5 57 | 45.4 77 | 42.4 10 | 36.5 18 | 55.5 23 | 50.3 62 |
Aniso. Huber-L1 [22] | 45.4 | 36.7 78 | 33.5 59 | 48.6 85 | 28.5 87 | 34.3 75 | 38.2 81 | 29.3 47 | 31.8 44 | 38.9 28 | 52.5 36 | 47.5 42 | 80.6 41 | 72.0 9 | 66.7 8 | 87.4 34 | 68.6 5 | 54.3 40 | 81.9 5 | 28.5 57 | 45.0 63 | 42.4 10 | 36.8 47 | 56.0 46 | 50.3 62 |
Classic+CPF [83] | 45.5 | 36.4 53 | 33.6 66 | 47.9 18 | 26.3 27 | 31.3 27 | 37.0 11 | 28.8 12 | 31.1 30 | 38.9 28 | 52.0 13 | 46.5 15 | 80.0 1 | 72.5 103 | 67.4 103 | 87.4 34 | 69.2 79 | 56.1 108 | 82.0 13 | 28.6 75 | 45.2 74 | 42.4 10 | 37.2 85 | 57.3 101 | 50.0 6 |
RFlow [90] | 46.8 | 36.2 28 | 33.0 43 | 48.2 57 | 27.6 68 | 33.7 66 | 37.1 18 | 29.3 47 | 32.5 57 | 39.2 62 | 52.6 49 | 47.8 56 | 80.6 41 | 72.0 9 | 66.8 17 | 87.3 16 | 68.6 5 | 53.8 15 | 81.9 5 | 28.5 57 | 45.5 83 | 42.6 90 | 37.2 85 | 56.9 87 | 50.3 62 |
EpicFlow [102] | 47.0 | 36.2 28 | 33.3 52 | 48.1 43 | 26.9 46 | 33.1 53 | 37.5 56 | 29.4 60 | 33.0 72 | 39.0 43 | 52.6 49 | 48.5 86 | 80.8 69 | 72.3 64 | 67.0 42 | 87.5 78 | 68.6 5 | 54.1 32 | 82.0 13 | 28.4 38 | 44.8 53 | 42.4 10 | 36.6 28 | 55.7 33 | 50.4 76 |
Brox et al. [5] | 47.8 | 36.3 42 | 32.4 23 | 48.2 57 | 27.8 74 | 34.1 72 | 38.0 78 | 29.8 82 | 33.9 92 | 39.6 88 | 52.5 36 | 47.0 29 | 80.4 19 | 72.2 36 | 66.9 29 | 87.5 78 | 68.7 19 | 53.8 15 | 82.1 35 | 28.4 38 | 44.9 56 | 42.5 62 | 36.5 18 | 55.5 23 | 50.2 45 |
S2D-Matching [84] | 48.5 | 36.6 70 | 34.2 91 | 48.2 57 | 26.9 46 | 32.5 42 | 37.2 31 | 28.8 12 | 30.7 19 | 38.9 28 | 52.1 20 | 46.4 11 | 80.9 77 | 72.3 64 | 67.1 65 | 87.5 78 | 69.1 69 | 55.3 87 | 82.2 57 | 28.5 57 | 44.7 44 | 42.4 10 | 36.7 40 | 55.9 43 | 50.2 45 |
FESL [72] | 49.2 | 36.6 70 | 33.9 80 | 48.0 29 | 26.4 29 | 31.7 30 | 37.3 36 | 29.1 30 | 31.3 33 | 38.9 28 | 52.6 49 | 47.6 46 | 80.3 10 | 72.4 91 | 67.3 93 | 87.4 34 | 69.3 90 | 55.9 103 | 82.1 35 | 28.4 38 | 44.9 56 | 42.3 1 | 37.0 68 | 56.6 78 | 50.1 24 |
p-harmonic [29] | 49.3 | 35.9 9 | 32.1 16 | 47.9 18 | 28.2 79 | 34.3 75 | 37.8 70 | 29.4 60 | 34.2 96 | 39.4 73 | 53.0 80 | 47.7 51 | 80.7 56 | 72.2 36 | 67.0 42 | 87.3 16 | 68.8 35 | 54.1 32 | 82.3 71 | 28.5 57 | 45.5 83 | 42.4 10 | 36.6 28 | 56.0 46 | 50.2 45 |
SepConv-v1 [127] | 49.8 | 27.1 1 | 27.5 1 | 36.4 1 | 24.9 1 | 31.0 20 | 40.3 94 | 27.6 1 | 28.4 1 | 48.9 128 | 54.0 104 | 47.3 35 | 83.0 119 | 72.0 9 | 66.7 8 | 87.1 3 | 69.1 69 | 52.6 1 | 83.6 124 | 32.2 130 | 43.9 16 | 55.7 130 | 37.0 68 | 53.8 1 | 55.4 130 |
ComplOF-FED-GPU [35] | 50.5 | 36.3 42 | 33.4 56 | 48.0 29 | 26.8 41 | 33.0 50 | 37.3 36 | 30.4 100 | 34.0 93 | 39.6 88 | 52.5 36 | 48.1 68 | 80.9 77 | 72.1 24 | 66.8 17 | 87.4 34 | 68.7 19 | 54.3 40 | 82.1 35 | 28.5 57 | 45.1 69 | 42.5 62 | 36.8 47 | 56.0 46 | 50.2 45 |
TC-Flow [46] | 51.3 | 36.2 28 | 33.2 49 | 48.2 57 | 26.9 46 | 33.5 63 | 37.5 56 | 29.5 70 | 33.6 86 | 38.9 28 | 52.1 20 | 47.1 32 | 80.6 41 | 72.3 64 | 67.2 79 | 87.5 78 | 69.0 59 | 54.8 66 | 82.3 71 | 28.4 38 | 44.4 32 | 42.5 62 | 36.6 28 | 56.1 54 | 50.1 24 |
EPPM w/o HM [88] | 51.6 | 35.8 4 | 32.3 22 | 47.6 3 | 26.7 36 | 33.0 50 | 36.9 3 | 30.0 89 | 35.5 111 | 39.4 73 | 52.6 49 | 48.9 100 | 80.4 19 | 72.2 36 | 67.1 65 | 87.4 34 | 69.3 90 | 55.9 103 | 82.3 71 | 28.4 38 | 44.9 56 | 42.5 62 | 36.8 47 | 56.1 54 | 50.1 24 |
DPOF [18] | 52.4 | 36.7 78 | 34.5 95 | 48.6 85 | 26.1 15 | 30.6 13 | 37.6 62 | 29.8 82 | 31.4 34 | 39.3 68 | 52.8 68 | 48.6 89 | 80.8 69 | 72.0 9 | 66.8 17 | 87.3 16 | 69.1 69 | 55.3 87 | 81.9 5 | 28.5 57 | 44.5 36 | 42.5 62 | 36.9 61 | 56.5 74 | 50.0 6 |
OFH [38] | 53.8 | 36.4 53 | 33.8 76 | 48.2 57 | 27.4 64 | 33.3 58 | 37.4 47 | 29.7 77 | 35.0 106 | 39.0 43 | 52.5 36 | 48.3 82 | 80.9 77 | 72.2 36 | 67.0 42 | 87.4 34 | 68.7 19 | 54.2 38 | 82.1 35 | 28.6 75 | 45.4 77 | 42.5 62 | 36.6 28 | 56.0 46 | 50.1 24 |
Efficient-NL [60] | 54.0 | 36.5 61 | 33.6 66 | 48.0 29 | 26.7 36 | 32.0 32 | 37.1 18 | 29.9 86 | 31.4 34 | 39.3 68 | 52.7 58 | 47.7 51 | 80.4 19 | 72.2 36 | 67.0 42 | 87.3 16 | 69.5 101 | 57.0 120 | 81.9 5 | 28.6 75 | 45.9 89 | 42.4 10 | 37.9 107 | 58.1 114 | 50.1 24 |
Local-TV-L1 [65] | 55.6 | 37.5 100 | 33.0 43 | 49.7 106 | 29.3 99 | 34.5 83 | 40.3 94 | 29.2 37 | 31.6 39 | 39.1 50 | 53.3 91 | 47.3 35 | 83.1 122 | 72.1 24 | 66.9 29 | 87.4 34 | 69.3 90 | 53.4 8 | 83.2 118 | 28.2 11 | 43.9 16 | 42.4 10 | 36.4 12 | 55.1 8 | 50.4 76 |
PMF [73] | 55.9 | 35.9 9 | 32.0 13 | 47.7 6 | 26.9 46 | 33.5 63 | 36.9 3 | 29.6 75 | 34.5 99 | 39.1 50 | 52.5 36 | 47.8 56 | 80.4 19 | 72.5 103 | 67.5 109 | 87.4 34 | 69.2 79 | 55.0 77 | 82.4 85 | 28.5 57 | 45.0 63 | 42.5 62 | 37.3 91 | 57.3 101 | 50.0 6 |
Sparse Occlusion [54] | 56.7 | 36.5 61 | 33.7 70 | 48.2 57 | 27.6 68 | 34.1 72 | 37.3 36 | 29.3 47 | 31.8 44 | 38.8 13 | 52.8 68 | 48.1 68 | 80.5 35 | 72.3 64 | 67.1 65 | 87.4 34 | 69.2 79 | 56.1 108 | 82.0 13 | 28.5 57 | 45.4 77 | 42.3 1 | 37.2 85 | 57.0 93 | 50.2 45 |
TC/T-Flow [76] | 56.8 | 36.6 70 | 33.7 70 | 47.9 18 | 26.8 41 | 32.9 47 | 37.1 18 | 29.1 30 | 31.9 49 | 38.8 13 | 52.7 58 | 48.5 86 | 80.4 19 | 72.5 103 | 67.4 103 | 87.5 78 | 69.1 69 | 55.2 82 | 82.1 35 | 28.6 75 | 45.4 77 | 42.5 62 | 37.0 68 | 56.9 87 | 50.0 6 |
OAR-Flow [125] | 56.8 | 36.5 61 | 33.0 43 | 48.4 75 | 27.0 52 | 33.0 50 | 37.8 70 | 29.2 37 | 33.3 80 | 38.9 28 | 52.1 20 | 47.6 46 | 80.5 35 | 72.4 91 | 67.3 93 | 87.6 110 | 68.9 45 | 54.5 49 | 82.2 57 | 28.5 57 | 44.7 44 | 42.4 10 | 36.9 61 | 56.5 74 | 50.4 76 |
TF+OM [100] | 57.2 | 36.3 42 | 33.0 43 | 48.5 78 | 26.9 46 | 32.2 35 | 39.2 89 | 28.6 3 | 32.4 55 | 38.9 28 | 52.8 68 | 48.2 80 | 80.7 56 | 72.3 64 | 67.1 65 | 87.4 34 | 69.0 59 | 54.5 49 | 82.3 71 | 28.4 38 | 45.1 69 | 42.5 62 | 37.0 68 | 56.4 71 | 50.6 100 |
SRR-TVOF-NL [91] | 57.5 | 36.6 70 | 33.5 59 | 48.2 57 | 27.7 70 | 34.3 75 | 37.9 74 | 29.5 70 | 33.2 76 | 39.1 50 | 53.1 84 | 48.1 68 | 80.2 5 | 72.2 36 | 67.1 65 | 87.3 16 | 68.9 45 | 55.7 99 | 81.8 3 | 28.5 57 | 44.9 56 | 42.4 10 | 37.5 98 | 58.0 113 | 50.1 24 |
SIOF [67] | 58.7 | 36.7 78 | 34.1 87 | 48.2 57 | 29.1 95 | 35.4 101 | 39.7 91 | 29.4 60 | 32.9 66 | 39.1 50 | 52.7 58 | 47.7 51 | 80.9 77 | 71.9 4 | 66.6 3 | 87.4 34 | 69.1 69 | 54.3 40 | 82.4 85 | 28.3 17 | 44.6 39 | 42.4 10 | 37.3 91 | 56.8 84 | 50.3 62 |
CLG-TV [48] | 59.0 | 36.6 70 | 33.4 56 | 48.5 78 | 28.2 79 | 34.4 80 | 38.2 81 | 29.7 77 | 33.6 86 | 39.4 73 | 52.8 68 | 48.0 62 | 80.9 77 | 72.2 36 | 66.9 29 | 87.5 78 | 68.7 19 | 54.0 28 | 82.1 35 | 28.4 38 | 45.1 69 | 42.4 10 | 37.0 68 | 56.5 74 | 50.2 45 |
ALD-Flow [66] | 59.2 | 36.7 78 | 33.9 80 | 48.6 85 | 27.0 52 | 33.2 55 | 37.9 74 | 29.3 47 | 33.4 83 | 38.9 28 | 52.5 36 | 48.0 62 | 80.9 77 | 72.4 91 | 67.2 79 | 87.6 110 | 68.9 45 | 54.4 45 | 82.2 57 | 28.2 11 | 43.6 8 | 42.4 10 | 37.0 68 | 56.6 78 | 50.3 62 |
SimpleFlow [49] | 59.8 | 36.5 61 | 34.2 91 | 48.2 57 | 27.2 60 | 32.8 46 | 37.3 36 | 30.1 95 | 31.7 41 | 39.4 73 | 52.0 13 | 46.3 10 | 80.7 56 | 72.3 64 | 67.2 79 | 87.4 34 | 69.0 59 | 55.4 93 | 82.0 13 | 28.7 86 | 47.1 103 | 42.6 90 | 37.0 68 | 56.8 84 | 50.1 24 |
Complementary OF [21] | 60.3 | 36.1 21 | 33.3 52 | 47.8 8 | 26.7 36 | 33.2 55 | 37.3 36 | 30.4 100 | 32.9 66 | 39.5 80 | 52.8 68 | 48.7 95 | 81.1 97 | 72.3 64 | 67.2 79 | 87.3 16 | 68.8 35 | 54.7 60 | 82.2 57 | 28.7 86 | 45.6 86 | 42.5 62 | 36.8 47 | 56.7 80 | 50.3 62 |
AggregFlow [97] | 60.6 | 37.1 93 | 34.8 103 | 48.5 78 | 27.3 62 | 33.2 55 | 38.1 79 | 28.7 9 | 30.2 9 | 38.5 2 | 52.9 76 | 48.6 89 | 80.3 10 | 72.4 91 | 67.2 79 | 87.6 110 | 69.3 90 | 54.5 49 | 82.6 102 | 28.3 17 | 44.2 27 | 42.5 62 | 36.7 40 | 56.0 46 | 50.4 76 |
LDOF [28] | 60.7 | 37.1 93 | 33.7 70 | 48.8 91 | 29.5 100 | 35.3 99 | 40.6 98 | 30.0 89 | 34.3 97 | 39.7 90 | 52.8 68 | 47.9 59 | 80.9 77 | 72.2 36 | 66.9 29 | 87.4 34 | 68.8 35 | 53.6 13 | 82.3 71 | 28.3 17 | 44.5 36 | 42.4 10 | 36.6 28 | 55.8 40 | 50.4 76 |
MLDP_OF [89] | 61.5 | 36.2 28 | 32.9 41 | 48.0 29 | 27.0 52 | 32.7 45 | 37.2 31 | 29.1 30 | 31.8 44 | 38.8 13 | 52.6 49 | 47.3 35 | 80.8 69 | 72.3 64 | 67.1 65 | 87.5 78 | 70.5 123 | 56.6 117 | 83.6 124 | 28.6 75 | 44.8 53 | 42.8 106 | 36.9 61 | 56.1 54 | 50.5 90 |
F-TV-L1 [15] | 61.6 | 37.4 98 | 34.6 97 | 49.2 97 | 28.8 93 | 34.9 93 | 38.3 83 | 29.7 77 | 34.1 95 | 39.5 80 | 52.7 58 | 47.6 46 | 81.0 89 | 71.7 2 | 66.5 2 | 87.4 34 | 68.8 35 | 53.5 12 | 82.4 85 | 28.3 17 | 44.3 30 | 42.4 10 | 37.1 78 | 56.3 67 | 50.6 100 |
IAOF [50] | 61.8 | 38.0 111 | 34.2 91 | 49.8 107 | 31.7 113 | 37.9 114 | 41.1 103 | 28.9 18 | 32.6 58 | 39.4 73 | 53.7 97 | 48.1 68 | 80.8 69 | 72.0 9 | 66.7 8 | 87.5 78 | 68.9 45 | 54.1 32 | 82.2 57 | 28.3 17 | 45.1 69 | 42.3 1 | 36.8 47 | 56.1 54 | 50.2 45 |
Aniso-Texture [82] | 62.3 | 36.1 21 | 32.4 23 | 48.2 57 | 28.0 77 | 34.4 80 | 37.6 62 | 30.0 89 | 32.8 65 | 39.3 68 | 52.7 58 | 48.1 68 | 80.7 56 | 72.4 91 | 67.3 93 | 87.3 16 | 69.2 79 | 56.2 114 | 82.3 71 | 28.4 38 | 44.5 36 | 42.4 10 | 37.2 85 | 57.0 93 | 50.2 45 |
Classic++ [32] | 63.2 | 36.4 53 | 33.5 59 | 48.4 75 | 27.4 64 | 33.7 66 | 37.6 62 | 29.6 75 | 33.6 86 | 39.2 62 | 52.7 58 | 47.3 35 | 80.9 77 | 72.2 36 | 67.0 42 | 87.5 78 | 69.1 69 | 54.5 49 | 82.5 96 | 28.5 57 | 44.9 56 | 42.6 90 | 36.8 47 | 56.2 62 | 50.3 62 |
Shiralkar [42] | 64.5 | 36.5 61 | 34.6 97 | 48.1 43 | 28.3 82 | 34.3 75 | 37.2 31 | 29.8 82 | 36.9 118 | 40.0 98 | 53.9 101 | 49.0 101 | 80.5 35 | 71.8 3 | 66.6 3 | 87.2 8 | 69.2 79 | 55.1 79 | 82.4 85 | 29.2 108 | 48.0 112 | 42.5 62 | 36.6 28 | 55.7 33 | 50.1 24 |
CostFilter [40] | 65.0 | 35.9 9 | 32.7 34 | 47.6 3 | 26.8 41 | 33.5 63 | 37.1 18 | 29.7 77 | 35.6 113 | 39.2 62 | 52.9 76 | 49.4 108 | 80.3 10 | 72.6 111 | 67.6 111 | 87.4 34 | 69.6 106 | 54.8 66 | 83.1 117 | 28.6 75 | 45.6 86 | 42.6 90 | 37.0 68 | 56.7 80 | 49.9 3 |
Fusion [6] | 65.2 | 36.0 16 | 32.7 34 | 47.8 8 | 26.8 41 | 32.1 34 | 37.5 56 | 29.5 70 | 31.5 38 | 39.5 80 | 53.5 93 | 48.6 89 | 80.7 56 | 72.6 111 | 68.0 119 | 87.1 3 | 69.3 90 | 57.6 126 | 81.8 3 | 28.7 86 | 47.1 103 | 42.5 62 | 38.2 114 | 59.9 126 | 50.0 6 |
FlowNetS+ft+v [112] | 65.2 | 36.8 84 | 33.0 43 | 48.7 88 | 29.5 100 | 35.6 102 | 40.5 96 | 29.8 82 | 34.3 97 | 39.5 80 | 52.8 68 | 48.2 80 | 80.8 69 | 72.2 36 | 67.0 42 | 87.4 34 | 68.7 19 | 53.9 23 | 82.1 35 | 28.6 75 | 45.9 89 | 42.5 62 | 36.7 40 | 56.0 46 | 50.4 76 |
TriFlow [95] | 65.5 | 37.0 90 | 35.3 109 | 48.8 91 | 28.7 91 | 34.5 83 | 41.0 101 | 29.2 37 | 33.4 83 | 38.8 13 | 53.0 80 | 48.8 99 | 80.4 19 | 72.3 64 | 67.3 93 | 87.4 34 | 69.2 79 | 55.5 95 | 82.1 35 | 28.5 57 | 44.8 53 | 42.4 10 | 36.9 61 | 56.4 71 | 50.1 24 |
SVFilterOh [111] | 65.6 | 36.3 42 | 32.2 18 | 48.1 43 | 26.2 22 | 30.9 17 | 37.4 47 | 29.2 37 | 30.6 16 | 39.3 68 | 52.6 49 | 47.5 42 | 81.0 89 | 72.6 111 | 67.6 111 | 87.6 110 | 69.3 90 | 55.9 103 | 82.3 71 | 28.5 57 | 43.7 13 | 43.3 120 | 37.3 91 | 57.1 97 | 51.0 110 |
Occlusion-TV-L1 [63] | 65.8 | 36.6 70 | 33.8 76 | 48.5 78 | 28.4 85 | 34.8 90 | 37.7 67 | 29.5 70 | 33.0 72 | 39.5 80 | 53.0 80 | 48.1 68 | 81.1 97 | 72.1 24 | 66.8 17 | 87.5 78 | 68.9 45 | 53.4 8 | 82.4 85 | 29.0 104 | 44.7 44 | 42.6 90 | 36.8 47 | 55.6 29 | 50.4 76 |
CNN-flow-warp+ref [117] | 66.3 | 36.3 42 | 31.7 9 | 48.7 88 | 28.5 87 | 34.7 86 | 39.5 90 | 30.4 100 | 35.0 106 | 39.8 93 | 54.0 104 | 48.1 68 | 81.2 101 | 72.3 64 | 67.0 42 | 87.4 34 | 68.6 5 | 53.2 5 | 82.4 85 | 28.8 94 | 47.1 103 | 42.5 62 | 36.6 28 | 55.7 33 | 50.3 62 |
CRTflow [80] | 66.5 | 36.7 78 | 33.8 76 | 48.5 78 | 27.7 70 | 33.8 69 | 37.4 47 | 30.7 106 | 35.3 108 | 40.9 115 | 52.9 76 | 48.1 68 | 81.8 111 | 72.2 36 | 66.9 29 | 87.4 34 | 68.9 45 | 54.1 32 | 82.3 71 | 28.4 38 | 44.9 56 | 42.5 62 | 36.8 47 | 56.1 54 | 50.5 90 |
FlowNet2 [122] | 68.5 | 39.4 116 | 38.2 120 | 50.4 113 | 29.2 98 | 34.8 90 | 41.9 110 | 30.0 89 | 34.6 101 | 39.4 73 | 53.3 91 | 51.0 120 | 80.6 41 | 72.5 103 | 67.4 103 | 87.4 34 | 68.8 35 | 54.3 40 | 82.0 13 | 28.4 38 | 45.0 63 | 42.3 1 | 36.5 18 | 55.7 33 | 49.8 1 |
Modified CLG [34] | 69.0 | 36.9 89 | 32.8 37 | 49.4 101 | 30.9 110 | 36.3 108 | 42.8 112 | 30.0 89 | 34.8 104 | 39.9 94 | 53.0 80 | 47.9 59 | 80.7 56 | 72.2 36 | 66.9 29 | 87.5 78 | 68.7 19 | 53.8 15 | 82.2 57 | 28.4 38 | 45.1 69 | 42.5 62 | 36.9 61 | 56.2 62 | 50.5 90 |
Adaptive [20] | 69.7 | 36.8 84 | 34.4 94 | 48.5 78 | 28.8 93 | 35.2 98 | 37.7 67 | 29.4 60 | 33.2 76 | 39.2 62 | 52.6 49 | 47.6 46 | 80.6 41 | 72.3 64 | 67.0 42 | 87.5 78 | 69.1 69 | 54.7 60 | 82.3 71 | 28.7 86 | 46.0 91 | 42.4 10 | 37.3 91 | 56.9 87 | 50.4 76 |
TCOF [69] | 69.8 | 36.6 70 | 33.9 80 | 48.1 43 | 29.1 95 | 35.7 103 | 38.3 83 | 29.0 22 | 31.4 34 | 38.7 8 | 52.8 68 | 48.7 95 | 80.6 41 | 72.2 36 | 67.1 65 | 87.4 34 | 69.3 90 | 56.0 106 | 82.1 35 | 28.7 86 | 46.2 94 | 42.5 62 | 38.2 114 | 58.7 122 | 50.5 90 |
Nguyen [33] | 71.8 | 39.6 117 | 33.9 80 | 52.6 121 | 32.5 118 | 37.9 114 | 43.3 115 | 30.0 89 | 35.5 111 | 40.2 103 | 54.1 107 | 49.0 101 | 80.9 77 | 72.0 9 | 66.8 17 | 87.4 34 | 68.6 5 | 53.8 15 | 82.0 13 | 28.8 94 | 47.8 110 | 42.4 10 | 36.8 47 | 56.1 54 | 50.3 62 |
Steered-L1 [118] | 72.0 | 36.0 16 | 32.9 41 | 47.9 18 | 27.0 52 | 33.3 58 | 37.7 67 | 30.3 99 | 32.3 52 | 39.9 94 | 53.2 86 | 48.0 62 | 81.0 89 | 72.5 103 | 67.5 109 | 87.5 78 | 68.9 45 | 55.0 77 | 82.2 57 | 28.8 94 | 46.7 100 | 42.7 101 | 37.0 68 | 57.3 101 | 50.3 62 |
StereoOF-V1MT [119] | 73.1 | 36.8 84 | 35.3 109 | 48.1 43 | 28.3 82 | 35.1 95 | 36.9 3 | 31.4 114 | 36.6 115 | 40.5 106 | 54.6 115 | 48.6 89 | 81.3 102 | 72.0 9 | 66.8 17 | 87.2 8 | 69.5 101 | 54.9 69 | 82.6 102 | 29.7 121 | 48.8 119 | 42.7 101 | 36.5 18 | 55.1 8 | 50.1 24 |
BriefMatch [124] | 73.1 | 36.3 42 | 33.3 52 | 48.0 29 | 27.2 60 | 33.4 61 | 38.5 87 | 30.6 105 | 32.6 58 | 40.6 108 | 54.0 104 | 48.6 89 | 82.8 118 | 72.4 91 | 67.3 93 | 87.3 16 | 70.2 118 | 55.6 98 | 83.9 127 | 28.3 17 | 44.3 30 | 42.7 101 | 36.6 28 | 55.7 33 | 50.5 90 |
SPSA-learn [13] | 76.5 | 37.4 98 | 33.6 66 | 49.4 101 | 29.8 104 | 35.1 95 | 41.4 107 | 30.9 109 | 33.2 76 | 40.7 109 | 53.5 93 | 47.2 33 | 80.4 19 | 72.2 36 | 67.0 42 | 87.4 34 | 68.8 35 | 54.1 32 | 82.2 57 | 29.5 116 | 52.2 130 | 42.9 111 | 37.1 78 | 57.0 93 | 50.3 62 |
GraphCuts [14] | 77.1 | 38.0 111 | 35.1 106 | 49.5 103 | 28.4 85 | 33.9 71 | 41.3 105 | 31.3 113 | 30.8 23 | 40.7 109 | 53.7 97 | 48.3 82 | 81.0 89 | 72.1 24 | 67.1 65 | 87.1 3 | 68.6 5 | 54.9 69 | 81.7 1 | 28.8 94 | 46.3 95 | 42.8 106 | 37.7 101 | 58.5 118 | 50.4 76 |
HBpMotionGpu [43] | 77.8 | 38.8 114 | 35.9 112 | 50.9 117 | 32.1 115 | 38.2 116 | 44.4 119 | 29.2 37 | 31.7 41 | 39.3 68 | 53.9 101 | 49.6 110 | 81.5 106 | 72.1 24 | 67.0 42 | 87.1 3 | 69.5 101 | 54.9 69 | 82.4 85 | 28.3 17 | 44.4 32 | 42.5 62 | 37.3 91 | 56.5 74 | 51.1 111 |
Dynamic MRF [7] | 78.0 | 36.2 28 | 34.1 87 | 48.0 29 | 27.5 66 | 34.6 85 | 37.4 47 | 30.9 109 | 36.8 117 | 40.4 105 | 54.5 113 | 49.3 107 | 81.9 112 | 71.9 4 | 66.8 17 | 87.2 8 | 69.4 99 | 55.5 95 | 82.5 96 | 29.0 104 | 47.8 110 | 42.5 62 | 37.5 98 | 56.8 84 | 50.5 90 |
2D-CLG [1] | 78.2 | 37.9 106 | 33.5 59 | 50.5 115 | 32.5 118 | 37.4 111 | 45.0 120 | 30.8 108 | 34.8 104 | 40.7 109 | 53.7 97 | 48.3 82 | 80.5 35 | 72.3 64 | 67.1 65 | 87.6 110 | 68.6 5 | 53.2 5 | 82.2 57 | 28.8 94 | 46.7 100 | 42.5 62 | 36.9 61 | 55.6 29 | 50.3 62 |
Black & Anandan [4] | 78.5 | 37.9 106 | 34.1 87 | 49.6 104 | 30.7 108 | 36.0 104 | 41.2 104 | 31.0 111 | 34.7 103 | 40.3 104 | 53.9 101 | 48.6 89 | 80.7 56 | 72.3 64 | 67.0 42 | 87.4 34 | 69.0 59 | 53.8 15 | 82.5 96 | 28.8 94 | 46.5 96 | 42.4 10 | 37.0 68 | 56.1 54 | 50.4 76 |
ROF-ND [107] | 78.6 | 37.0 90 | 32.8 37 | 48.1 43 | 27.7 70 | 34.7 86 | 37.6 62 | 29.7 77 | 32.3 52 | 39.1 50 | 54.2 109 | 51.4 121 | 80.4 19 | 72.4 91 | 67.3 93 | 87.4 34 | 69.5 101 | 56.9 118 | 82.0 13 | 29.7 121 | 49.0 122 | 43.2 118 | 37.8 105 | 57.7 110 | 50.2 45 |
TV-L1-improved [17] | 79.2 | 36.6 70 | 34.1 87 | 48.4 75 | 28.6 90 | 35.1 95 | 37.8 70 | 30.5 103 | 33.2 76 | 40.0 98 | 52.7 58 | 48.0 62 | 80.9 77 | 72.3 64 | 67.2 79 | 87.4 34 | 69.1 69 | 54.9 69 | 82.3 71 | 28.8 94 | 47.3 106 | 42.6 90 | 37.2 85 | 56.7 80 | 50.6 100 |
CBF [12] | 80.9 | 36.4 53 | 32.5 29 | 48.9 94 | 27.5 66 | 33.8 69 | 37.9 74 | 29.3 47 | 31.6 39 | 39.1 50 | 53.2 86 | 48.1 68 | 82.6 115 | 72.4 91 | 67.2 79 | 87.7 119 | 69.2 79 | 55.3 87 | 82.3 71 | 28.7 86 | 46.1 92 | 42.9 111 | 37.9 107 | 57.7 110 | 51.7 120 |
UnFlow [129] | 82.4 | 39.2 115 | 37.9 118 | 50.6 116 | 32.3 116 | 38.9 121 | 41.3 105 | 31.6 119 | 38.5 123 | 40.8 114 | 53.2 86 | 48.7 95 | 80.9 77 | 72.0 9 | 66.7 8 | 87.4 34 | 69.5 101 | 54.6 56 | 82.4 85 | 28.2 11 | 43.2 3 | 42.4 10 | 39.3 127 | 58.3 116 | 51.2 112 |
Rannacher [23] | 82.4 | 36.7 78 | 34.5 95 | 48.7 88 | 28.7 91 | 35.3 99 | 38.1 79 | 30.5 103 | 34.0 93 | 39.9 94 | 52.7 58 | 48.0 62 | 80.8 69 | 72.4 91 | 67.2 79 | 87.5 78 | 69.0 59 | 54.6 56 | 82.3 71 | 28.8 94 | 47.0 102 | 42.6 90 | 37.1 78 | 56.4 71 | 50.6 100 |
Correlation Flow [75] | 82.9 | 36.2 28 | 33.4 56 | 47.7 6 | 27.7 70 | 34.3 75 | 37.3 36 | 29.4 60 | 31.4 34 | 38.8 13 | 53.1 84 | 48.5 86 | 81.3 102 | 72.8 116 | 67.6 111 | 88.6 129 | 70.1 113 | 57.1 121 | 82.6 102 | 29.4 113 | 48.8 119 | 43.0 113 | 37.7 101 | 57.9 112 | 50.5 90 |
HBM-GC [105] | 84.0 | 37.7 103 | 34.7 100 | 49.8 107 | 27.1 57 | 32.4 40 | 37.9 74 | 28.8 12 | 29.6 3 | 39.2 62 | 52.6 49 | 47.3 35 | 80.8 69 | 73.2 122 | 68.1 120 | 88.2 124 | 70.0 111 | 57.3 123 | 82.7 106 | 28.9 103 | 45.0 63 | 43.5 121 | 37.6 100 | 57.2 99 | 51.3 114 |
SegOF [10] | 84.6 | 37.6 102 | 33.2 49 | 50.0 110 | 29.1 95 | 34.7 86 | 41.0 101 | 31.4 114 | 35.3 108 | 40.7 109 | 53.6 95 | 50.7 118 | 80.6 41 | 72.3 64 | 67.2 79 | 87.5 78 | 69.0 59 | 55.3 87 | 82.2 57 | 29.0 104 | 48.6 117 | 42.7 101 | 36.7 40 | 55.8 40 | 50.4 76 |
TriangleFlow [30] | 84.7 | 37.0 90 | 34.9 104 | 48.5 78 | 28.0 77 | 34.7 86 | 37.5 56 | 30.2 97 | 33.0 72 | 39.9 94 | 53.2 86 | 49.0 101 | 81.1 97 | 72.0 9 | 66.9 29 | 87.1 3 | 69.8 110 | 56.1 108 | 82.4 85 | 29.2 108 | 48.5 116 | 42.8 106 | 38.1 112 | 58.5 118 | 50.5 90 |
BlockOverlap [61] | 85.1 | 38.5 113 | 33.2 49 | 51.3 118 | 30.0 106 | 34.4 80 | 42.8 112 | 29.4 60 | 30.4 14 | 40.0 98 | 53.2 86 | 46.9 25 | 83.0 119 | 72.9 119 | 67.6 111 | 88.3 125 | 69.7 108 | 54.1 32 | 83.3 122 | 28.7 86 | 44.1 25 | 43.5 121 | 37.1 78 | 55.3 14 | 51.8 121 |
IAOF2 [51] | 87.4 | 37.9 106 | 35.9 112 | 49.1 96 | 29.6 102 | 36.1 106 | 40.0 93 | 29.3 47 | 33.4 83 | 40.0 98 | 54.1 107 | 50.2 115 | 81.0 89 | 72.4 91 | 67.4 103 | 87.4 34 | 69.2 79 | 54.9 69 | 82.4 85 | 28.6 75 | 45.5 83 | 42.4 10 | 37.9 107 | 57.6 107 | 50.6 100 |
Ad-TV-NDC [36] | 87.5 | 40.4 121 | 35.1 106 | 53.1 122 | 31.9 114 | 36.7 110 | 43.8 117 | 29.4 60 | 32.9 66 | 39.1 50 | 54.5 113 | 49.2 106 | 82.1 113 | 72.5 103 | 67.3 93 | 87.5 78 | 69.3 90 | 53.9 23 | 82.7 106 | 28.6 75 | 45.4 77 | 42.4 10 | 37.2 85 | 56.2 62 | 50.6 100 |
AdaConv-v1 [126] | 93.9 | 37.2 97 | 36.6 115 | 47.5 2 | 34.3 123 | 39.1 123 | 51.1 129 | 36.1 127 | 39.4 125 | 52.9 131 | 58.2 127 | 53.1 125 | 83.8 124 | 70.9 1 | 65.4 1 | 86.6 2 | 69.7 108 | 54.7 60 | 84.4 128 | 38.6 131 | 46.5 96 | 77.4 131 | 38.2 114 | 54.6 2 | 60.3 131 |
LocallyOriented [52] | 95.4 | 37.5 100 | 35.9 112 | 49.2 97 | 29.6 102 | 36.2 107 | 39.1 88 | 30.1 95 | 33.8 90 | 39.5 80 | 53.7 97 | 50.0 113 | 81.3 102 | 72.3 64 | 67.2 79 | 87.5 78 | 70.2 118 | 56.2 114 | 82.9 113 | 28.8 94 | 45.6 86 | 42.5 62 | 37.7 101 | 57.6 107 | 50.5 90 |
ACK-Prior [27] | 96.4 | 36.4 53 | 33.7 70 | 48.1 43 | 26.7 36 | 33.1 53 | 37.1 18 | 30.7 106 | 33.3 80 | 39.7 90 | 53.6 95 | 50.0 113 | 81.0 89 | 73.5 126 | 68.6 123 | 88.3 125 | 70.8 125 | 59.8 129 | 82.7 106 | 29.7 121 | 48.7 118 | 43.6 124 | 39.5 128 | 62.1 129 | 51.3 114 |
Horn & Schunck [3] | 96.6 | 37.9 106 | 35.1 106 | 49.6 104 | 31.4 111 | 37.7 113 | 41.8 109 | 31.7 120 | 37.4 120 | 41.5 117 | 55.8 119 | 50.6 116 | 81.3 102 | 72.2 36 | 67.0 42 | 87.4 34 | 69.2 79 | 54.2 38 | 82.7 106 | 29.5 116 | 48.9 121 | 42.6 90 | 37.8 105 | 57.2 99 | 50.9 109 |
StereoFlow [44] | 96.7 | 46.3 130 | 45.9 131 | 54.3 123 | 38.3 130 | 45.4 131 | 45.7 122 | 29.3 47 | 33.8 90 | 39.1 50 | 52.9 76 | 47.7 51 | 81.0 89 | 74.4 130 | 70.5 131 | 87.6 110 | 72.0 129 | 66.3 131 | 82.4 85 | 28.4 38 | 45.0 63 | 42.4 10 | 38.0 111 | 59.1 123 | 50.5 90 |
Filter Flow [19] | 97.8 | 37.8 105 | 34.6 97 | 49.8 107 | 30.8 109 | 36.0 104 | 44.3 118 | 29.4 60 | 32.4 55 | 39.5 80 | 54.2 109 | 48.1 68 | 82.2 114 | 72.7 115 | 67.7 117 | 87.6 110 | 69.2 79 | 55.1 79 | 82.5 96 | 28.7 86 | 46.5 96 | 42.6 90 | 38.3 120 | 58.4 117 | 51.4 117 |
TI-DOFE [24] | 98.7 | 42.0 124 | 37.5 117 | 54.8 126 | 35.2 125 | 41.1 128 | 46.8 125 | 31.4 114 | 37.7 121 | 41.6 119 | 56.1 121 | 50.6 116 | 81.6 107 | 72.0 9 | 66.9 29 | 87.2 8 | 69.4 99 | 54.4 45 | 82.6 102 | 29.2 108 | 47.6 109 | 42.6 90 | 38.2 114 | 57.5 105 | 50.8 108 |
SILK [79] | 100.3 | 39.6 117 | 38.1 119 | 51.5 120 | 32.4 117 | 38.5 119 | 43.6 116 | 32.4 121 | 37.2 119 | 41.5 117 | 55.4 117 | 49.7 111 | 83.0 119 | 72.2 36 | 67.0 42 | 87.4 34 | 70.0 111 | 54.7 60 | 83.4 123 | 29.0 104 | 46.5 96 | 42.8 106 | 37.4 97 | 56.7 80 | 50.7 106 |
Bartels [41] | 102.0 | 37.1 93 | 35.0 105 | 49.3 100 | 28.2 79 | 34.8 90 | 40.5 96 | 29.9 86 | 33.1 75 | 40.5 106 | 54.2 109 | 49.7 111 | 83.9 125 | 73.0 120 | 67.6 111 | 88.7 130 | 71.8 128 | 56.1 108 | 85.6 130 | 28.6 75 | 43.9 16 | 43.6 124 | 38.1 112 | 57.0 93 | 53.2 127 |
SLK [47] | 108.6 | 41.6 123 | 38.7 122 | 54.4 124 | 33.0 120 | 38.3 118 | 45.5 121 | 33.3 122 | 38.6 124 | 42.8 121 | 57.8 125 | 51.8 123 | 83.5 123 | 72.1 24 | 67.3 93 | 86.5 1 | 70.1 113 | 55.8 101 | 82.7 106 | 30.0 124 | 51.4 128 | 43.0 113 | 38.2 114 | 57.5 105 | 51.5 118 |
NL-TV-NCC [25] | 109.9 | 37.1 93 | 35.7 111 | 48.0 29 | 27.8 74 | 35.0 94 | 37.6 62 | 31.0 111 | 35.4 110 | 40.0 98 | 56.0 120 | 54.2 128 | 82.6 115 | 73.8 128 | 68.6 123 | 89.1 131 | 70.6 124 | 58.4 128 | 82.5 96 | 30.4 128 | 50.0 125 | 44.0 128 | 39.8 129 | 60.2 127 | 52.4 125 |
GroupFlow [9] | 110.6 | 40.3 120 | 40.1 124 | 51.3 118 | 31.5 112 | 38.9 121 | 42.6 111 | 33.5 124 | 39.5 126 | 43.8 123 | 54.7 116 | 52.3 124 | 81.0 89 | 73.2 122 | 68.6 123 | 87.6 110 | 70.4 122 | 57.3 123 | 83.0 115 | 29.3 111 | 48.1 113 | 42.5 62 | 37.9 107 | 58.2 115 | 50.1 24 |
FFV1MT [106] | 111.2 | 39.6 117 | 40.8 126 | 50.3 112 | 34.8 124 | 38.8 120 | 46.6 124 | 36.5 128 | 45.8 129 | 44.6 125 | 56.2 122 | 49.0 101 | 81.7 108 | 72.5 103 | 67.4 103 | 87.4 34 | 70.2 118 | 54.7 60 | 83.2 118 | 30.1 126 | 49.3 124 | 42.8 106 | 38.6 121 | 57.6 107 | 51.3 114 |
Learning Flow [11] | 111.6 | 37.7 103 | 37.0 116 | 49.2 97 | 29.9 105 | 37.5 112 | 39.7 91 | 31.5 118 | 36.3 114 | 40.7 109 | 55.4 117 | 51.6 122 | 82.6 115 | 72.8 116 | 67.8 118 | 87.8 121 | 69.6 106 | 55.7 99 | 82.8 111 | 29.3 111 | 48.4 114 | 42.7 101 | 39.2 125 | 59.8 125 | 51.2 112 |
Heeger++ [104] | 111.7 | 40.6 122 | 42.5 128 | 50.4 113 | 33.5 121 | 38.2 116 | 43.0 114 | 37.6 129 | 48.1 130 | 44.9 126 | 56.2 122 | 49.0 101 | 81.7 108 | 73.4 125 | 68.9 128 | 87.5 78 | 70.1 113 | 56.0 106 | 82.8 111 | 30.3 127 | 49.1 123 | 42.6 90 | 37.7 101 | 56.9 87 | 50.3 62 |
2bit-BM-tele [98] | 112.5 | 37.9 106 | 34.7 100 | 50.1 111 | 30.1 107 | 36.4 109 | 41.7 108 | 30.2 97 | 32.2 51 | 41.3 116 | 54.3 112 | 49.4 108 | 84.1 127 | 73.3 124 | 68.1 120 | 88.3 125 | 72.2 130 | 57.2 122 | 85.5 129 | 30.4 128 | 52.7 131 | 44.4 129 | 38.2 114 | 56.3 67 | 54.1 129 |
FOLKI [16] | 116.3 | 44.6 128 | 40.4 125 | 58.4 129 | 35.7 126 | 42.3 129 | 47.3 126 | 33.3 122 | 40.7 127 | 44.9 126 | 59.4 130 | 53.6 126 | 86.5 130 | 72.5 103 | 67.6 111 | 87.3 16 | 70.1 113 | 55.8 101 | 83.2 118 | 29.4 113 | 48.4 114 | 43.1 116 | 38.7 122 | 58.5 118 | 52.0 122 |
Pyramid LK [2] | 120.0 | 46.1 129 | 38.9 123 | 61.0 130 | 36.7 129 | 40.4 126 | 50.9 128 | 39.9 130 | 36.6 115 | 49.4 129 | 64.1 131 | 61.2 131 | 87.7 131 | 73.1 121 | 68.6 123 | 87.4 34 | 70.1 113 | 56.1 108 | 83.0 115 | 29.6 119 | 50.7 127 | 43.2 118 | 39.2 125 | 61.2 128 | 51.5 118 |
Adaptive flow [45] | 120.8 | 43.8 126 | 38.2 120 | 56.5 127 | 35.8 127 | 40.5 127 | 50.2 127 | 31.4 114 | 34.5 99 | 42.5 120 | 56.5 124 | 50.9 119 | 83.9 125 | 73.5 126 | 68.7 127 | 88.1 123 | 70.2 118 | 57.4 125 | 82.9 113 | 29.4 113 | 47.5 107 | 43.6 124 | 39.0 124 | 59.1 123 | 52.0 122 |
PGAM+LK [55] | 121.4 | 42.5 125 | 41.4 127 | 54.6 125 | 33.7 122 | 40.1 125 | 45.8 123 | 33.9 125 | 41.1 128 | 43.3 122 | 59.3 129 | 55.2 129 | 85.5 129 | 72.8 116 | 68.1 120 | 87.5 78 | 70.8 125 | 56.9 118 | 83.6 124 | 29.6 119 | 50.0 125 | 43.0 113 | 38.7 122 | 58.6 121 | 52.1 124 |
HCIC-L [99] | 125.0 | 49.1 131 | 42.6 129 | 63.0 131 | 35.8 127 | 39.4 124 | 52.5 130 | 34.6 126 | 37.7 121 | 43.9 124 | 58.0 126 | 53.9 127 | 81.7 108 | 74.0 129 | 69.3 129 | 88.5 128 | 71.7 127 | 60.5 130 | 83.2 118 | 29.5 116 | 47.5 107 | 43.9 127 | 40.5 130 | 62.9 130 | 52.8 126 |
Periodicity [78] | 128.8 | 44.4 127 | 43.3 130 | 56.9 128 | 42.8 131 | 43.4 130 | 56.2 131 | 40.9 131 | 49.1 131 | 49.5 130 | 58.9 128 | 58.6 130 | 84.9 128 | 74.4 130 | 70.2 130 | 88.0 122 | 73.1 131 | 57.9 127 | 86.2 131 | 30.0 124 | 51.5 129 | 43.5 121 | 41.8 131 | 63.4 131 | 53.7 128 |
Method | time* | frames | color | 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 | C. Bailer, B. Taetz, and D. Stricker. Flow Fields: Dense unregularized correspondence fields for highly accurate large displacement optical flow estimation. ICCV 2015. | |
[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. | |
[130] 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. | |
[131] Kuang | 9.9 | 2 | gray | F. Kuang. PatchMatch algorithms for motion estimation and stereo reconstruction. Master thesis, University of Stuttgart, 2017. |