Optical flow evaluation results |
Statistics:
Average
SD
R2.5
R5.0
R10.0
A90
A95
A99
Error type: endpoint angle interpolation normalized interpolation |
R5.0 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] | 12.4 | 4.93 3 | 13.9 3 | 0.13 4 | 8.97 27 | 17.1 12 | 0.43 10 | 6.00 10 | 13.4 4 | 0.27 1 | 17.6 3 | 26.2 6 | 5.24 20 | 43.0 14 | 57.7 8 | 5.17 20 | 10.3 5 | 39.1 8 | 0.87 12 | 9.75 17 | 41.0 16 | 0.44 17 | 21.5 28 | 51.9 35 | 0.47 15 |
MDP-Flow2 [68] | 13.4 | 4.89 2 | 14.4 5 | 0.12 3 | 8.58 11 | 16.9 10 | 0.39 3 | 5.95 7 | 13.6 6 | 0.28 3 | 17.7 4 | 26.7 14 | 5.32 34 | 42.9 6 | 57.6 5 | 5.13 15 | 10.6 19 | 40.1 25 | 0.92 22 | 9.75 17 | 41.0 16 | 0.43 10 | 21.6 40 | 51.9 35 | 0.46 9 |
SepConv-v1 [127] | 14.9 | 3.41 1 | 11.0 1 | 0.08 1 | 8.39 7 | 16.7 9 | 1.04 99 | 2.81 1 | 7.63 1 | 0.74 99 | 18.0 17 | 25.2 1 | 5.82 98 | 42.9 6 | 57.4 2 | 4.74 2 | 9.03 1 | 34.1 1 | 0.60 1 | 9.34 2 | 38.6 1 | 0.42 4 | 20.1 1 | 48.6 1 | 0.35 1 |
NN-field [71] | 17.9 | 5.14 19 | 16.1 48 | 0.13 4 | 8.21 3 | 15.7 3 | 0.38 2 | 6.39 43 | 13.6 6 | 0.30 7 | 18.4 35 | 28.7 70 | 5.33 38 | 42.9 6 | 57.6 5 | 5.08 11 | 10.7 26 | 40.2 30 | 0.94 29 | 9.62 7 | 40.5 7 | 0.44 17 | 21.2 4 | 51.2 7 | 0.45 3 |
NNF-Local [87] | 18.0 | 5.11 15 | 15.7 26 | 0.11 2 | 8.18 2 | 15.8 4 | 0.39 3 | 6.01 13 | 13.5 5 | 0.27 1 | 18.3 33 | 28.3 54 | 5.29 27 | 43.0 14 | 57.6 5 | 5.11 13 | 10.8 42 | 40.9 56 | 1.01 46 | 9.67 9 | 40.7 9 | 0.46 33 | 21.2 4 | 51.2 7 | 0.46 9 |
Layers++ [37] | 21.5 | 5.25 32 | 15.9 34 | 0.17 35 | 8.27 4 | 15.5 2 | 0.37 1 | 6.16 22 | 14.3 12 | 0.38 39 | 18.0 17 | 26.9 18 | 5.32 34 | 43.1 25 | 57.9 21 | 5.24 37 | 10.7 26 | 40.6 48 | 0.97 41 | 9.70 10 | 40.7 9 | 0.39 1 | 21.3 10 | 51.3 9 | 0.48 28 |
PH-Flow [101] | 24.3 | 5.32 44 | 16.4 59 | 0.16 25 | 8.28 5 | 15.9 5 | 0.44 13 | 6.12 19 | 13.9 9 | 0.33 16 | 17.5 1 | 25.8 2 | 5.15 7 | 42.8 4 | 57.5 3 | 5.03 8 | 11.0 67 | 41.6 85 | 1.09 68 | 9.71 11 | 41.0 16 | 0.46 33 | 21.3 10 | 51.4 11 | 0.50 63 |
nLayers [57] | 26.0 | 5.26 35 | 15.8 32 | 0.16 25 | 8.54 9 | 16.6 8 | 0.45 16 | 5.89 4 | 13.1 2 | 0.30 7 | 18.1 23 | 27.1 23 | 5.35 44 | 43.3 50 | 58.0 30 | 5.36 69 | 10.8 42 | 40.9 56 | 1.11 71 | 9.65 8 | 40.1 3 | 0.48 56 | 21.2 4 | 51.1 5 | 0.45 3 |
COFM [59] | 26.1 | 5.08 13 | 15.1 13 | 0.19 53 | 8.86 19 | 17.4 17 | 0.48 25 | 6.37 39 | 14.2 11 | 0.40 47 | 17.7 4 | 26.2 6 | 5.11 2 | 42.9 6 | 57.8 12 | 5.02 7 | 10.9 53 | 41.6 85 | 1.11 71 | 9.24 1 | 38.8 2 | 0.50 69 | 21.5 28 | 51.9 35 | 0.46 9 |
Sparse-NonSparse [56] | 26.8 | 5.31 43 | 16.3 57 | 0.17 35 | 8.74 14 | 17.2 16 | 0.48 25 | 6.19 23 | 14.7 24 | 0.34 20 | 17.9 12 | 26.3 8 | 5.23 17 | 43.1 25 | 57.8 12 | 5.25 40 | 11.0 67 | 41.2 65 | 1.04 54 | 9.71 11 | 40.9 13 | 0.46 33 | 21.2 4 | 51.3 9 | 0.47 15 |
IROF++ [58] | 28.0 | 5.37 56 | 16.8 75 | 0.14 8 | 8.87 21 | 17.4 17 | 0.45 16 | 6.41 49 | 14.6 21 | 0.43 57 | 17.5 1 | 25.8 2 | 5.22 12 | 42.9 6 | 57.8 12 | 5.19 25 | 10.5 11 | 39.4 13 | 0.87 12 | 10.0 55 | 42.4 61 | 0.47 49 | 21.4 19 | 51.5 12 | 0.50 63 |
TV-L1-MCT [64] | 28.5 | 5.74 103 | 18.1 109 | 0.18 45 | 9.50 41 | 19.1 40 | 0.58 40 | 5.73 2 | 14.5 19 | 0.38 39 | 17.8 8 | 26.0 5 | 5.28 25 | 43.0 14 | 57.9 21 | 5.22 31 | 10.4 7 | 39.1 8 | 0.94 29 | 9.78 23 | 41.1 22 | 0.44 17 | 21.2 4 | 51.1 5 | 0.48 28 |
HAST [109] | 29.7 | 5.12 16 | 15.2 15 | 0.16 25 | 8.74 14 | 17.1 12 | 0.43 10 | 6.62 74 | 15.3 46 | 0.39 44 | 17.7 4 | 26.4 10 | 4.98 1 | 43.0 14 | 58.0 30 | 5.05 9 | 11.0 67 | 41.4 74 | 1.06 61 | 9.53 3 | 40.4 5 | 0.42 4 | 22.0 81 | 52.8 78 | 0.47 15 |
ComponentFusion [96] | 30.8 | 5.15 20 | 16.1 48 | 0.14 8 | 8.86 19 | 17.9 25 | 0.41 6 | 6.38 40 | 15.4 47 | 0.33 16 | 17.8 8 | 27.0 21 | 5.15 7 | 43.2 40 | 58.0 30 | 5.24 37 | 10.6 19 | 39.8 17 | 0.94 29 | 10.0 55 | 42.7 80 | 0.57 96 | 21.5 28 | 51.8 28 | 0.47 15 |
ProbFlowFields [128] | 31.2 | 5.03 8 | 15.6 24 | 0.17 35 | 8.55 10 | 17.1 12 | 0.41 6 | 6.00 10 | 14.4 15 | 0.32 13 | 18.1 23 | 27.1 23 | 5.38 49 | 43.3 50 | 58.1 46 | 5.49 105 | 10.9 53 | 41.2 65 | 1.20 89 | 9.61 6 | 40.7 9 | 0.47 49 | 21.0 3 | 50.8 3 | 0.49 44 |
FMOF [94] | 31.5 | 5.62 92 | 17.2 87 | 0.21 62 | 8.71 13 | 17.0 11 | 0.44 13 | 6.38 40 | 14.7 24 | 0.46 62 | 18.6 48 | 28.0 41 | 5.31 32 | 43.1 25 | 57.9 21 | 5.15 18 | 10.8 42 | 40.5 44 | 0.87 12 | 9.60 5 | 40.4 5 | 0.40 3 | 21.5 28 | 51.7 20 | 0.46 9 |
2DHMM-SAS [92] | 34.7 | 5.62 92 | 17.6 103 | 0.18 45 | 10.1 64 | 19.7 55 | 0.64 55 | 5.73 2 | 14.4 15 | 0.37 37 | 17.7 4 | 25.9 4 | 5.30 29 | 43.0 14 | 57.8 12 | 5.26 43 | 10.7 26 | 40.0 23 | 0.82 5 | 9.83 29 | 41.3 25 | 0.48 56 | 21.6 40 | 52.0 39 | 0.47 15 |
CombBMOF [113] | 35.2 | 5.46 70 | 16.2 55 | 0.22 72 | 8.89 22 | 18.0 27 | 0.45 16 | 6.29 30 | 14.7 24 | 0.40 47 | 18.5 42 | 28.0 41 | 5.24 20 | 43.0 14 | 57.7 8 | 5.08 11 | 10.8 42 | 40.2 30 | 0.82 5 | 11.7 125 | 42.9 85 | 0.47 49 | 21.2 4 | 50.9 4 | 0.45 3 |
LSM [39] | 36.2 | 5.49 74 | 17.4 97 | 0.18 45 | 8.93 24 | 17.7 22 | 0.48 25 | 6.32 34 | 15.4 47 | 0.35 28 | 18.1 23 | 27.1 23 | 5.22 12 | 43.1 25 | 57.9 21 | 5.28 53 | 11.0 67 | 41.3 70 | 1.03 52 | 9.72 13 | 40.9 13 | 0.46 33 | 21.4 19 | 51.7 20 | 0.48 28 |
Ramp [62] | 37.1 | 5.46 70 | 17.1 84 | 0.18 45 | 8.84 17 | 17.4 17 | 0.58 40 | 6.14 20 | 14.7 24 | 0.34 20 | 17.8 8 | 26.4 10 | 5.23 17 | 43.2 40 | 58.0 30 | 5.27 48 | 11.2 87 | 42.0 93 | 1.15 79 | 9.72 13 | 40.9 13 | 0.42 4 | 21.6 40 | 52.1 44 | 0.48 28 |
DeepFlow [86] | 37.6 | 5.06 12 | 14.6 7 | 0.19 53 | 9.80 56 | 19.5 46 | 0.75 67 | 6.45 52 | 16.6 81 | 0.35 28 | 18.7 57 | 27.6 34 | 5.41 55 | 43.4 67 | 58.0 30 | 5.37 71 | 10.3 5 | 38.3 4 | 0.99 42 | 9.83 29 | 41.8 41 | 0.43 10 | 21.3 10 | 51.6 18 | 0.48 28 |
SuperFlow [81] | 38.0 | 4.99 7 | 14.3 4 | 0.22 72 | 10.3 70 | 19.9 60 | 0.90 82 | 6.61 69 | 15.5 50 | 0.51 69 | 18.5 42 | 27.2 28 | 5.52 74 | 43.3 50 | 58.1 46 | 5.37 71 | 10.1 3 | 38.0 3 | 0.73 3 | 9.73 16 | 41.4 30 | 0.46 33 | 21.3 10 | 51.5 12 | 0.46 9 |
NNF-EAC [103] | 38.1 | 5.52 78 | 15.7 26 | 0.34 113 | 9.27 37 | 18.1 29 | 0.48 25 | 6.53 58 | 13.8 8 | 0.40 47 | 18.2 29 | 27.0 21 | 5.71 87 | 43.0 14 | 57.7 8 | 5.11 13 | 10.4 7 | 39.1 8 | 0.83 7 | 9.89 37 | 41.6 36 | 0.52 80 | 21.7 51 | 52.2 51 | 0.49 44 |
DeepFlow2 [108] | 38.2 | 5.16 21 | 14.9 12 | 0.21 62 | 9.81 57 | 19.7 55 | 0.65 57 | 6.38 40 | 16.3 70 | 0.34 20 | 18.6 48 | 28.1 46 | 5.29 27 | 43.4 67 | 58.0 30 | 5.37 71 | 10.2 4 | 38.4 5 | 0.85 10 | 9.96 52 | 42.1 54 | 0.44 17 | 21.4 19 | 51.8 28 | 0.49 44 |
LME [70] | 38.4 | 5.13 18 | 15.8 32 | 0.14 8 | 9.15 34 | 18.4 38 | 0.51 31 | 6.32 34 | 15.7 54 | 0.34 20 | 17.9 12 | 27.1 23 | 5.34 40 | 43.8 110 | 58.8 109 | 5.79 123 | 10.8 42 | 41.2 65 | 0.93 24 | 9.86 33 | 41.3 25 | 0.43 10 | 21.3 10 | 51.5 12 | 0.47 15 |
WLIF-Flow [93] | 39.2 | 5.25 32 | 16.0 41 | 0.15 15 | 9.14 33 | 18.1 29 | 0.59 46 | 6.29 30 | 14.3 12 | 0.34 20 | 17.9 12 | 26.3 8 | 5.65 83 | 43.1 25 | 57.9 21 | 5.26 43 | 11.2 87 | 41.9 92 | 1.22 95 | 9.82 28 | 41.3 25 | 0.44 17 | 21.7 51 | 52.2 51 | 0.49 44 |
PGM-C [120] | 39.2 | 5.18 24 | 16.0 41 | 0.15 15 | 8.97 27 | 18.2 31 | 0.46 22 | 6.51 55 | 16.4 75 | 0.33 16 | 18.4 35 | 28.5 61 | 5.36 46 | 43.4 67 | 58.1 46 | 5.40 87 | 10.7 26 | 40.5 44 | 0.96 38 | 9.92 42 | 41.9 43 | 0.45 25 | 21.4 19 | 51.8 28 | 0.48 28 |
FlowFields+ [130] | 39.3 | 5.23 31 | 16.6 67 | 0.15 15 | 8.91 23 | 18.3 32 | 0.45 16 | 6.28 29 | 15.9 58 | 0.34 20 | 18.2 29 | 28.1 46 | 5.34 40 | 43.4 67 | 58.2 57 | 5.35 66 | 10.9 53 | 41.6 85 | 1.10 70 | 9.79 24 | 41.5 32 | 0.46 33 | 21.3 10 | 51.5 12 | 0.48 28 |
FlowFields [110] | 40.0 | 5.22 29 | 16.5 62 | 0.16 25 | 8.95 25 | 18.3 32 | 0.42 8 | 6.29 30 | 15.9 58 | 0.35 28 | 18.4 35 | 28.5 61 | 5.41 55 | 43.4 67 | 58.1 46 | 5.33 58 | 10.9 53 | 41.3 70 | 1.08 64 | 9.79 24 | 41.5 32 | 0.45 25 | 21.3 10 | 51.6 18 | 0.49 44 |
Classic+NL [31] | 40.4 | 5.56 84 | 17.4 97 | 0.22 72 | 8.99 29 | 17.6 21 | 0.54 33 | 6.02 14 | 14.7 24 | 0.36 33 | 18.1 23 | 26.8 15 | 5.41 55 | 43.1 25 | 58.0 30 | 5.23 34 | 11.1 84 | 41.5 79 | 1.06 61 | 9.72 13 | 41.0 16 | 0.46 33 | 21.6 40 | 52.0 39 | 0.47 15 |
DF-Auto [115] | 41.4 | 5.03 8 | 13.8 2 | 0.17 35 | 10.2 65 | 19.3 42 | 0.79 71 | 6.09 16 | 14.4 15 | 0.34 20 | 18.7 57 | 28.1 46 | 5.24 20 | 43.2 40 | 57.9 21 | 5.31 56 | 10.4 7 | 39.3 11 | 0.93 24 | 10.1 64 | 42.3 58 | 0.49 62 | 21.9 73 | 52.9 84 | 0.53 96 |
S2F-IF [123] | 42.3 | 5.22 29 | 16.5 62 | 0.15 15 | 8.84 17 | 18.0 27 | 0.44 13 | 6.27 28 | 15.7 54 | 0.33 16 | 18.3 33 | 28.3 54 | 5.14 6 | 43.4 67 | 58.2 57 | 5.41 90 | 11.0 67 | 41.5 79 | 1.11 71 | 9.91 41 | 41.9 43 | 0.47 49 | 21.3 10 | 51.5 12 | 0.51 75 |
FC-2Layers-FF [74] | 42.4 | 5.40 61 | 17.0 81 | 0.17 35 | 8.15 1 | 15.3 1 | 0.42 8 | 6.14 20 | 14.9 32 | 0.35 28 | 18.1 23 | 27.2 28 | 5.31 32 | 43.3 50 | 58.2 57 | 5.36 69 | 11.2 87 | 42.2 97 | 1.20 89 | 9.75 17 | 41.0 16 | 0.49 62 | 21.7 51 | 52.1 44 | 0.48 28 |
AGIF+OF [85] | 42.6 | 5.60 89 | 17.4 97 | 0.15 15 | 8.95 25 | 17.7 22 | 0.59 46 | 6.20 25 | 14.5 19 | 0.43 57 | 17.9 12 | 26.6 13 | 5.22 12 | 43.4 67 | 58.3 79 | 5.38 78 | 11.1 84 | 42.0 93 | 1.01 46 | 9.87 36 | 40.7 9 | 0.42 4 | 21.5 28 | 52.0 39 | 0.48 28 |
OFLAF [77] | 42.9 | 5.16 21 | 15.9 34 | 0.14 8 | 8.28 5 | 16.1 6 | 0.40 5 | 6.34 38 | 14.9 32 | 0.30 7 | 18.0 17 | 27.3 30 | 5.11 2 | 43.3 50 | 58.1 46 | 5.39 80 | 11.2 87 | 42.4 98 | 1.21 92 | 10.1 64 | 42.4 61 | 0.60 102 | 21.9 73 | 52.6 69 | 0.45 3 |
MDP-Flow [26] | 43.8 | 5.03 8 | 15.4 17 | 0.14 8 | 8.68 12 | 17.4 17 | 0.47 23 | 5.97 8 | 14.3 12 | 0.32 13 | 18.9 73 | 28.5 61 | 5.50 71 | 43.2 40 | 58.0 30 | 5.39 80 | 11.2 87 | 42.6 101 | 1.31 103 | 10.3 82 | 43.1 90 | 0.49 62 | 21.4 19 | 51.7 20 | 0.47 15 |
S2D-Matching [84] | 44.8 | 5.56 84 | 17.3 91 | 0.18 45 | 9.96 61 | 19.9 60 | 0.66 58 | 5.99 9 | 14.7 24 | 0.41 52 | 17.9 12 | 26.4 10 | 5.40 54 | 43.2 40 | 58.0 30 | 5.17 20 | 11.2 87 | 42.0 93 | 1.17 84 | 9.93 45 | 41.1 22 | 0.43 10 | 21.5 28 | 51.8 28 | 0.48 28 |
TF+OM [100] | 45.6 | 4.98 5 | 14.6 7 | 0.20 58 | 9.03 31 | 17.9 25 | 0.55 35 | 6.29 30 | 16.2 67 | 0.39 44 | 18.5 42 | 28.0 41 | 5.50 71 | 43.3 50 | 58.1 46 | 5.47 102 | 10.6 19 | 39.8 17 | 1.03 52 | 9.86 33 | 42.0 49 | 0.51 76 | 21.7 51 | 52.3 56 | 0.52 87 |
Brox et al. [5] | 46.4 | 5.33 50 | 15.4 17 | 0.19 53 | 10.2 65 | 20.1 65 | 0.64 55 | 6.61 69 | 17.2 96 | 0.46 62 | 18.7 57 | 28.2 50 | 5.21 9 | 43.4 67 | 58.1 46 | 5.27 48 | 10.7 26 | 40.1 25 | 0.99 42 | 9.90 39 | 42.0 49 | 0.45 25 | 21.6 40 | 52.1 44 | 0.47 15 |
ALD-Flow [66] | 46.4 | 5.37 56 | 16.1 48 | 0.23 80 | 9.53 43 | 19.2 41 | 0.57 38 | 6.51 55 | 16.7 85 | 0.34 20 | 18.2 29 | 27.9 37 | 5.32 34 | 43.4 67 | 58.3 79 | 5.46 100 | 10.7 26 | 39.9 20 | 0.99 42 | 9.76 22 | 41.2 24 | 0.44 17 | 21.8 62 | 52.7 74 | 0.47 15 |
CPM-Flow [116] | 47.1 | 5.20 28 | 16.1 48 | 0.16 25 | 8.99 29 | 18.3 32 | 0.47 23 | 6.42 50 | 16.0 63 | 0.30 7 | 18.8 65 | 29.2 89 | 5.43 62 | 43.4 67 | 58.2 57 | 5.44 98 | 10.6 19 | 40.1 25 | 1.02 48 | 10.0 55 | 42.6 72 | 0.45 25 | 21.4 19 | 51.8 28 | 0.53 96 |
SVFilterOh [111] | 47.2 | 5.32 44 | 15.7 26 | 0.21 62 | 8.78 16 | 17.1 12 | 0.49 29 | 6.40 46 | 14.6 21 | 0.38 39 | 18.4 35 | 27.1 23 | 5.80 96 | 43.8 110 | 58.6 105 | 5.65 117 | 10.9 53 | 41.0 62 | 1.04 54 | 9.54 4 | 40.1 3 | 0.43 10 | 21.7 51 | 52.2 51 | 0.50 63 |
AggregFlow [97] | 47.6 | 5.64 95 | 17.2 87 | 0.22 72 | 9.81 57 | 19.5 46 | 0.59 46 | 6.11 18 | 14.4 15 | 0.28 3 | 18.9 73 | 29.0 80 | 5.30 29 | 43.4 67 | 58.2 57 | 5.33 58 | 10.7 26 | 40.2 30 | 0.96 38 | 9.89 37 | 41.7 38 | 0.50 69 | 21.4 19 | 51.7 20 | 0.50 63 |
RNLOD-Flow [121] | 48.9 | 5.32 44 | 16.6 67 | 0.16 25 | 9.70 52 | 19.6 52 | 0.60 50 | 6.57 62 | 15.5 50 | 0.51 69 | 18.2 29 | 27.4 31 | 5.22 12 | 43.1 25 | 58.0 30 | 5.28 53 | 11.0 67 | 41.4 74 | 1.08 64 | 9.85 32 | 41.3 25 | 0.50 69 | 21.9 73 | 52.7 74 | 0.49 44 |
Second-order prior [8] | 49.1 | 5.29 40 | 15.3 16 | 0.27 97 | 10.8 79 | 21.1 79 | 0.78 70 | 7.14 94 | 17.8 103 | 0.62 92 | 18.6 48 | 28.3 54 | 5.21 9 | 42.9 6 | 57.7 8 | 5.16 19 | 10.5 11 | 39.6 15 | 0.93 24 | 10.2 75 | 42.8 82 | 0.44 17 | 21.6 40 | 52.3 56 | 0.49 44 |
IROF-TV [53] | 50.5 | 5.35 55 | 16.6 67 | 0.21 62 | 9.10 32 | 17.8 24 | 0.57 38 | 6.61 69 | 16.8 87 | 0.44 59 | 17.8 8 | 26.9 18 | 5.37 48 | 43.5 90 | 58.4 91 | 5.50 107 | 10.5 11 | 40.1 25 | 0.90 18 | 9.98 54 | 42.2 56 | 0.46 33 | 21.6 40 | 52.1 44 | 0.51 75 |
DPOF [18] | 51.2 | 5.51 77 | 17.9 107 | 0.22 72 | 8.45 8 | 16.5 7 | 0.43 10 | 6.87 81 | 15.1 41 | 0.59 84 | 18.9 73 | 29.5 93 | 5.43 62 | 42.9 6 | 57.8 12 | 5.05 9 | 11.0 67 | 40.9 56 | 0.84 9 | 10.3 82 | 42.5 68 | 0.45 25 | 21.9 73 | 52.8 78 | 0.48 28 |
TC-Flow [46] | 52.9 | 5.19 25 | 15.9 34 | 0.21 62 | 9.57 44 | 19.6 52 | 0.63 52 | 6.78 79 | 17.0 93 | 0.36 33 | 18.1 23 | 27.4 31 | 5.61 79 | 43.3 50 | 58.2 57 | 5.46 100 | 11.0 67 | 41.6 85 | 1.18 85 | 9.93 45 | 41.7 38 | 0.45 25 | 21.5 28 | 52.0 39 | 0.49 44 |
Aniso. Huber-L1 [22] | 53.1 | 5.41 63 | 16.0 41 | 0.23 80 | 11.2 89 | 21.1 79 | 0.90 82 | 6.72 76 | 15.4 47 | 0.46 62 | 18.5 42 | 28.1 46 | 5.39 53 | 43.0 14 | 57.8 12 | 5.23 34 | 10.5 11 | 40.1 25 | 0.81 4 | 10.2 75 | 42.6 72 | 0.46 33 | 21.9 73 | 52.7 74 | 0.52 87 |
Kuang [131] | 53.4 | 5.34 52 | 17.0 81 | 0.16 25 | 9.50 41 | 19.4 45 | 0.49 29 | 6.58 65 | 16.4 75 | 0.38 39 | 18.6 48 | 28.9 75 | 5.42 59 | 43.4 67 | 58.2 57 | 5.39 80 | 10.7 26 | 40.7 51 | 0.90 18 | 10.3 82 | 43.5 96 | 0.52 80 | 21.4 19 | 51.8 28 | 0.49 44 |
OAR-Flow [125] | 53.5 | 5.28 38 | 15.5 20 | 0.18 45 | 9.71 54 | 19.5 46 | 0.67 59 | 6.43 51 | 16.3 70 | 0.28 3 | 18.0 17 | 27.6 34 | 5.23 17 | 43.5 90 | 58.4 91 | 5.48 103 | 10.9 53 | 41.3 70 | 1.13 76 | 10.2 75 | 42.9 85 | 0.51 76 | 21.7 51 | 52.3 56 | 0.45 3 |
ComplOF-FED-GPU [35] | 54.2 | 5.30 42 | 16.1 48 | 0.19 53 | 9.39 39 | 19.3 42 | 0.58 40 | 7.21 98 | 16.9 90 | 0.66 94 | 18.4 35 | 28.6 67 | 5.32 34 | 43.1 25 | 58.0 30 | 5.27 48 | 10.8 42 | 40.9 56 | 0.99 42 | 10.1 64 | 42.8 82 | 0.47 49 | 21.8 62 | 52.3 56 | 0.50 63 |
EpicFlow [102] | 54.2 | 5.19 25 | 16.1 48 | 0.15 15 | 9.60 45 | 19.8 59 | 0.58 40 | 6.40 46 | 16.4 75 | 0.35 28 | 18.6 48 | 29.1 87 | 5.47 68 | 43.4 67 | 58.2 57 | 5.42 93 | 10.8 42 | 41.2 65 | 1.08 64 | 10.1 64 | 42.5 68 | 0.54 87 | 21.5 28 | 52.0 39 | 0.49 44 |
FESL [72] | 56.0 | 5.65 98 | 17.3 91 | 0.17 35 | 9.18 35 | 18.3 32 | 0.55 35 | 6.22 26 | 15.0 38 | 0.44 59 | 18.8 65 | 28.4 57 | 5.38 49 | 43.4 67 | 58.2 57 | 5.41 90 | 11.3 94 | 42.8 105 | 1.19 87 | 9.92 42 | 41.5 32 | 0.42 4 | 21.8 62 | 52.3 56 | 0.48 28 |
Classic+CPF [83] | 56.2 | 5.59 88 | 17.3 91 | 0.16 25 | 9.22 36 | 18.3 32 | 0.58 40 | 6.00 10 | 14.9 32 | 0.40 47 | 18.0 17 | 26.8 15 | 5.22 12 | 43.5 90 | 58.5 99 | 5.38 78 | 11.4 99 | 43.0 111 | 1.15 79 | 10.1 64 | 41.9 43 | 0.45 25 | 22.0 81 | 53.1 90 | 0.49 44 |
PMF [73] | 56.6 | 5.32 44 | 16.6 67 | 0.14 8 | 9.67 51 | 19.9 60 | 0.45 16 | 6.89 86 | 18.2 107 | 0.49 66 | 18.4 35 | 27.9 37 | 5.21 9 | 43.5 90 | 58.4 91 | 5.22 31 | 11.0 67 | 40.5 44 | 1.27 100 | 9.86 33 | 41.8 41 | 0.46 33 | 22.1 89 | 53.1 90 | 0.50 63 |
RFlow [90] | 57.8 | 5.19 25 | 16.1 48 | 0.23 80 | 10.8 79 | 21.2 83 | 0.85 77 | 6.59 67 | 16.0 63 | 0.51 69 | 18.8 65 | 28.8 73 | 5.47 68 | 43.1 25 | 58.0 30 | 5.21 29 | 10.5 11 | 40.0 23 | 0.93 24 | 10.0 55 | 42.6 72 | 0.49 62 | 22.1 89 | 53.2 94 | 0.51 75 |
Local-TV-L1 [65] | 57.8 | 5.29 40 | 14.6 7 | 0.35 115 | 11.5 96 | 21.1 79 | 1.23 107 | 6.39 43 | 14.9 32 | 0.37 37 | 19.0 79 | 27.9 37 | 6.64 112 | 43.3 50 | 58.3 79 | 5.33 58 | 10.9 53 | 39.0 6 | 1.58 123 | 9.79 24 | 41.6 36 | 0.48 56 | 21.3 10 | 51.5 12 | 0.53 96 |
CLG-TV [48] | 59.0 | 5.32 44 | 15.7 26 | 0.26 94 | 11.0 85 | 21.2 83 | 0.83 76 | 6.75 78 | 16.6 81 | 0.56 78 | 18.9 73 | 28.4 57 | 5.50 71 | 43.3 50 | 58.1 46 | 5.25 40 | 10.5 11 | 39.8 17 | 0.87 12 | 10.1 64 | 42.5 68 | 0.44 17 | 22.0 81 | 53.1 90 | 0.51 75 |
TriFlow [95] | 60.0 | 5.42 64 | 17.0 81 | 0.24 86 | 10.9 82 | 21.2 83 | 0.91 84 | 6.61 69 | 16.8 87 | 0.36 33 | 18.9 73 | 29.0 80 | 5.28 25 | 43.2 40 | 58.2 57 | 5.37 71 | 11.0 67 | 40.9 56 | 0.95 33 | 9.96 52 | 41.7 38 | 0.49 62 | 21.7 51 | 52.2 51 | 0.47 15 |
Classic++ [32] | 60.0 | 5.33 50 | 16.0 41 | 0.28 98 | 10.2 65 | 20.3 69 | 0.69 62 | 6.87 81 | 16.6 81 | 0.50 67 | 18.7 57 | 27.7 36 | 5.64 81 | 43.2 40 | 58.0 30 | 5.26 43 | 11.0 67 | 40.7 51 | 1.34 106 | 9.93 45 | 41.9 43 | 0.47 49 | 21.7 51 | 52.4 65 | 0.50 63 |
EPPM w/o HM [88] | 60.1 | 5.34 52 | 17.3 91 | 0.13 4 | 9.73 55 | 20.1 65 | 0.53 32 | 7.33 105 | 18.7 113 | 0.63 93 | 18.5 42 | 29.1 87 | 5.33 38 | 43.1 25 | 58.0 30 | 5.20 28 | 11.0 67 | 41.4 74 | 0.96 38 | 10.3 82 | 42.3 58 | 0.56 93 | 21.8 62 | 52.4 65 | 0.49 44 |
SIOF [67] | 60.5 | 5.64 95 | 16.5 62 | 0.28 98 | 11.3 91 | 21.6 90 | 0.91 84 | 6.32 34 | 15.9 58 | 0.42 53 | 18.7 57 | 28.4 57 | 5.36 46 | 43.0 14 | 57.9 21 | 5.17 20 | 10.7 26 | 40.2 30 | 0.95 33 | 10.1 64 | 42.4 61 | 0.50 69 | 22.2 98 | 53.2 94 | 0.53 96 |
LDOF [28] | 61.1 | 5.53 82 | 15.6 24 | 0.32 110 | 11.1 87 | 20.3 69 | 1.45 121 | 6.89 86 | 17.3 98 | 0.59 84 | 19.0 79 | 28.9 75 | 5.63 80 | 43.4 67 | 58.2 57 | 5.40 87 | 10.4 7 | 39.0 6 | 0.83 7 | 9.92 42 | 42.4 61 | 0.46 33 | 21.6 40 | 52.3 56 | 0.46 9 |
Efficient-NL [60] | 61.5 | 5.54 83 | 17.1 84 | 0.16 25 | 9.60 45 | 18.9 39 | 0.56 37 | 6.99 90 | 15.1 41 | 0.75 100 | 18.8 65 | 28.2 50 | 5.26 24 | 43.1 25 | 57.9 21 | 5.25 40 | 11.6 105 | 43.4 119 | 1.04 54 | 10.1 64 | 42.5 68 | 0.48 56 | 22.6 108 | 53.8 106 | 0.48 28 |
p-harmonic [29] | 61.8 | 5.17 23 | 15.5 20 | 0.16 25 | 11.2 89 | 21.4 87 | 0.94 87 | 6.55 59 | 17.4 101 | 0.55 77 | 19.2 87 | 28.6 67 | 5.45 66 | 43.3 50 | 58.2 57 | 5.27 48 | 10.7 26 | 40.2 30 | 1.04 54 | 10.4 91 | 43.4 94 | 0.50 69 | 21.8 62 | 52.6 69 | 0.49 44 |
Complementary OF [21] | 62.0 | 5.28 38 | 16.7 73 | 0.15 15 | 9.39 39 | 19.5 46 | 0.58 40 | 7.53 109 | 16.3 70 | 1.10 118 | 18.7 57 | 29.0 80 | 5.35 44 | 43.2 40 | 58.2 57 | 5.26 43 | 10.9 53 | 41.2 65 | 1.16 82 | 10.3 82 | 43.4 94 | 0.55 90 | 21.5 28 | 52.2 51 | 0.51 75 |
F-TV-L1 [15] | 62.2 | 5.56 84 | 16.0 41 | 0.36 118 | 11.4 94 | 21.5 89 | 0.94 87 | 6.88 83 | 17.0 93 | 0.66 94 | 18.7 57 | 27.9 37 | 5.79 95 | 42.6 2 | 57.8 12 | 5.01 6 | 10.6 19 | 39.3 11 | 1.02 48 | 10.0 55 | 41.9 43 | 0.55 90 | 22.0 81 | 52.8 78 | 0.51 75 |
CostFilter [40] | 62.6 | 5.44 66 | 17.7 104 | 0.13 4 | 9.64 48 | 20.1 65 | 0.45 16 | 6.96 89 | 19.1 115 | 0.47 65 | 18.5 42 | 28.9 75 | 5.13 5 | 43.6 104 | 58.5 99 | 5.32 57 | 11.1 84 | 40.5 44 | 1.48 117 | 9.94 49 | 42.1 54 | 0.45 25 | 21.8 62 | 52.6 69 | 0.49 44 |
OFH [38] | 63.2 | 5.49 74 | 16.6 67 | 0.25 90 | 10.3 70 | 20.2 68 | 0.77 69 | 6.88 83 | 17.8 103 | 0.36 33 | 18.4 35 | 28.9 75 | 5.24 20 | 43.1 25 | 58.0 30 | 5.26 43 | 10.9 53 | 41.5 79 | 1.18 85 | 10.3 82 | 43.0 88 | 0.58 98 | 21.6 40 | 52.1 44 | 0.50 63 |
CBF [12] | 63.7 | 4.98 5 | 14.8 11 | 0.18 45 | 10.2 65 | 19.9 60 | 0.71 64 | 6.63 75 | 15.2 45 | 0.42 53 | 19.0 79 | 28.5 61 | 6.39 109 | 43.4 67 | 58.3 79 | 5.49 105 | 10.7 26 | 40.4 40 | 0.95 33 | 10.1 64 | 42.6 72 | 0.50 69 | 22.3 102 | 53.5 103 | 0.53 96 |
HBM-GC [105] | 63.7 | 5.52 78 | 17.1 84 | 0.22 72 | 9.64 48 | 19.3 42 | 0.59 46 | 5.93 6 | 13.2 3 | 0.31 12 | 18.8 65 | 28.0 41 | 5.83 100 | 44.3 121 | 59.2 115 | 5.71 119 | 11.5 102 | 43.3 117 | 1.32 104 | 9.75 17 | 40.6 8 | 0.39 1 | 22.0 81 | 52.9 84 | 0.50 63 |
TC/T-Flow [76] | 63.8 | 5.73 101 | 17.3 91 | 0.22 72 | 9.66 50 | 19.7 55 | 0.63 52 | 6.24 27 | 14.9 32 | 0.32 13 | 18.6 48 | 28.7 70 | 5.38 49 | 43.5 90 | 58.4 91 | 5.50 107 | 11.0 67 | 41.4 74 | 0.89 17 | 10.2 75 | 43.0 88 | 0.58 98 | 21.9 73 | 53.0 89 | 0.45 3 |
Steered-L1 [118] | 63.9 | 5.12 16 | 16.0 41 | 0.17 35 | 9.62 47 | 19.5 46 | 0.88 79 | 7.15 95 | 15.6 52 | 1.00 110 | 19.4 91 | 28.5 61 | 6.39 109 | 43.5 90 | 58.5 99 | 5.19 25 | 10.8 42 | 40.8 55 | 1.20 89 | 9.95 51 | 42.6 72 | 0.52 80 | 21.7 51 | 52.6 69 | 0.48 28 |
GraphCuts [14] | 64.5 | 5.98 111 | 17.5 101 | 0.24 86 | 10.0 62 | 19.5 46 | 0.76 68 | 8.24 120 | 14.6 21 | 1.06 113 | 19.7 94 | 29.0 80 | 5.69 85 | 42.9 6 | 57.9 21 | 4.97 4 | 10.5 11 | 40.3 35 | 0.87 12 | 10.0 55 | 42.4 61 | 0.58 98 | 22.1 89 | 53.2 94 | 0.51 75 |
AdaConv-v1 [126] | 64.7 | 6.72 120 | 21.8 124 | 0.25 90 | 12.8 112 | 22.4 107 | 1.80 126 | 8.18 119 | 18.4 109 | 1.46 125 | 24.3 125 | 34.7 126 | 7.39 120 | 41.5 1 | 56.1 1 | 4.28 1 | 9.57 2 | 36.9 2 | 0.71 2 | 9.75 17 | 41.0 16 | 0.60 102 | 20.5 2 | 49.7 2 | 0.42 2 |
MLDP_OF [89] | 66.1 | 5.44 66 | 17.2 87 | 0.17 35 | 9.84 59 | 19.9 60 | 0.62 51 | 6.19 23 | 14.8 30 | 0.28 3 | 18.6 48 | 27.4 31 | 5.71 87 | 43.3 50 | 58.2 57 | 5.34 63 | 11.9 114 | 43.3 117 | 1.57 122 | 10.4 91 | 42.6 72 | 0.56 93 | 21.7 51 | 52.3 56 | 0.59 120 |
BlockOverlap [61] | 66.2 | 5.34 52 | 14.6 7 | 0.41 122 | 11.4 94 | 20.6 72 | 1.42 117 | 6.49 53 | 14.1 10 | 0.61 90 | 18.9 73 | 26.9 18 | 7.34 119 | 44.2 119 | 58.9 111 | 5.91 125 | 11.0 67 | 39.9 20 | 1.39 112 | 9.81 27 | 41.3 25 | 0.46 33 | 21.5 28 | 51.7 20 | 0.51 75 |
SRR-TVOF-NL [91] | 66.8 | 5.70 99 | 16.9 79 | 0.23 80 | 10.3 70 | 21.0 78 | 0.88 79 | 6.57 62 | 16.1 65 | 0.39 44 | 19.2 87 | 28.7 70 | 5.12 4 | 43.2 40 | 58.3 79 | 5.27 48 | 10.8 42 | 40.9 56 | 0.86 11 | 10.6 101 | 42.3 58 | 0.46 33 | 22.5 104 | 53.8 106 | 0.54 107 |
Sparse Occlusion [54] | 68.0 | 5.43 65 | 16.8 75 | 0.23 80 | 10.3 70 | 20.8 76 | 0.63 52 | 6.51 55 | 15.0 38 | 0.44 59 | 19.0 79 | 29.0 80 | 5.42 59 | 43.4 67 | 58.2 57 | 5.41 90 | 11.3 94 | 42.9 109 | 1.14 77 | 10.1 64 | 42.2 56 | 0.42 4 | 22.1 89 | 53.2 94 | 0.49 44 |
CRTflow [80] | 68.5 | 5.48 73 | 16.5 62 | 0.34 113 | 10.7 77 | 20.7 73 | 0.86 78 | 7.25 100 | 18.6 112 | 0.60 89 | 18.8 65 | 28.8 73 | 5.98 103 | 43.4 67 | 58.2 57 | 5.43 95 | 10.7 26 | 40.4 40 | 0.95 33 | 9.93 45 | 42.0 49 | 0.49 62 | 21.7 51 | 52.3 56 | 0.49 44 |
SimpleFlow [49] | 70.5 | 5.52 78 | 17.5 101 | 0.18 45 | 10.2 65 | 19.7 55 | 0.73 65 | 7.32 104 | 15.8 56 | 1.05 112 | 18.0 17 | 26.8 15 | 5.44 64 | 43.3 50 | 58.1 46 | 5.33 58 | 11.3 94 | 42.9 109 | 1.22 95 | 10.3 82 | 44.6 107 | 1.04 127 | 21.8 62 | 52.6 69 | 0.47 15 |
FlowNet2 [122] | 70.7 | 6.90 121 | 21.5 123 | 0.25 90 | 10.6 76 | 20.7 73 | 0.82 73 | 7.10 93 | 17.3 98 | 0.54 74 | 19.4 91 | 31.8 117 | 5.57 76 | 43.4 67 | 58.3 79 | 5.39 80 | 10.7 26 | 40.3 35 | 0.90 18 | 10.0 55 | 42.0 49 | 0.46 33 | 21.6 40 | 51.9 35 | 0.51 75 |
IAOF [50] | 71.3 | 5.97 110 | 16.8 75 | 0.29 103 | 14.1 125 | 24.8 125 | 1.41 116 | 6.05 15 | 16.2 67 | 0.61 90 | 20.1 101 | 29.5 93 | 5.47 68 | 43.0 14 | 57.8 12 | 5.19 25 | 10.7 26 | 40.3 35 | 0.94 29 | 10.4 91 | 43.3 92 | 0.46 33 | 22.0 81 | 52.8 78 | 0.54 107 |
Modified CLG [34] | 71.6 | 5.05 11 | 15.1 13 | 0.19 53 | 12.3 109 | 22.2 102 | 1.30 110 | 6.81 80 | 18.3 108 | 0.66 94 | 19.3 90 | 29.7 96 | 5.34 40 | 43.4 67 | 58.2 57 | 5.29 55 | 10.8 42 | 40.6 48 | 1.15 79 | 10.2 75 | 43.6 97 | 0.47 49 | 21.9 73 | 52.7 74 | 0.53 96 |
Aniso-Texture [82] | 72.8 | 5.09 14 | 15.7 26 | 0.15 15 | 11.1 87 | 21.7 91 | 1.00 92 | 7.30 103 | 15.9 58 | 0.59 84 | 18.7 57 | 28.6 67 | 5.90 101 | 43.6 104 | 58.4 91 | 5.53 110 | 11.6 105 | 44.0 124 | 1.44 115 | 9.90 39 | 41.4 30 | 0.43 10 | 22.1 89 | 53.1 90 | 0.49 44 |
FlowNetS+ft+v [112] | 73.3 | 5.40 61 | 15.5 20 | 0.29 103 | 11.7 102 | 21.7 91 | 1.62 123 | 6.88 83 | 17.1 95 | 0.56 78 | 19.0 79 | 29.2 89 | 5.73 91 | 43.5 90 | 58.4 91 | 5.56 112 | 10.5 11 | 39.9 20 | 0.95 33 | 10.1 64 | 42.9 85 | 0.52 80 | 21.8 62 | 52.5 68 | 0.48 28 |
Occlusion-TV-L1 [63] | 74.2 | 5.32 44 | 16.2 55 | 0.28 98 | 11.3 91 | 21.9 96 | 0.96 91 | 6.60 68 | 16.9 90 | 0.58 82 | 19.1 85 | 28.9 75 | 5.72 89 | 43.4 67 | 58.2 57 | 5.24 37 | 10.9 53 | 40.3 35 | 1.26 99 | 10.9 108 | 42.6 72 | 0.81 119 | 21.8 62 | 52.4 65 | 0.49 44 |
Shiralkar [42] | 75.2 | 5.73 101 | 18.1 109 | 0.21 62 | 11.6 98 | 22.0 98 | 0.88 79 | 6.74 77 | 19.9 119 | 0.73 98 | 20.3 104 | 30.1 101 | 5.46 67 | 42.6 2 | 57.5 3 | 4.99 5 | 11.3 94 | 41.5 79 | 1.35 107 | 11.0 110 | 44.9 109 | 0.67 107 | 21.5 28 | 51.7 20 | 0.48 28 |
CNN-flow-warp+ref [117] | 76.0 | 4.95 4 | 14.4 5 | 0.22 72 | 10.9 82 | 21.2 83 | 1.23 107 | 7.43 107 | 18.0 105 | 0.79 102 | 20.9 112 | 29.8 97 | 6.84 115 | 43.5 90 | 58.3 79 | 5.57 113 | 10.7 26 | 40.3 35 | 1.22 95 | 10.3 82 | 44.4 106 | 0.67 107 | 21.6 40 | 52.1 44 | 0.47 15 |
TCOF [69] | 76.2 | 5.56 84 | 16.8 75 | 0.17 35 | 11.8 103 | 22.1 100 | 1.02 96 | 6.09 16 | 15.0 38 | 0.30 7 | 19.0 79 | 29.4 92 | 5.67 84 | 43.4 67 | 58.3 79 | 5.17 20 | 11.4 99 | 43.1 113 | 1.02 48 | 11.0 110 | 43.9 99 | 0.48 56 | 23.1 119 | 55.1 123 | 0.52 87 |
HBpMotionGpu [43] | 76.3 | 5.80 105 | 16.3 57 | 0.42 123 | 13.1 115 | 23.8 117 | 1.34 112 | 6.32 34 | 14.9 32 | 0.38 39 | 19.9 95 | 30.4 104 | 5.80 96 | 43.1 25 | 58.3 79 | 5.39 80 | 11.3 94 | 41.0 62 | 1.21 92 | 9.94 49 | 41.9 43 | 0.43 10 | 22.1 89 | 52.9 84 | 0.53 96 |
Fusion [6] | 77.3 | 5.37 56 | 16.9 79 | 0.21 62 | 9.33 38 | 18.3 32 | 0.54 33 | 6.39 43 | 15.1 41 | 0.54 74 | 20.0 99 | 29.8 97 | 5.41 55 | 43.5 90 | 59.2 115 | 5.14 16 | 11.5 102 | 43.7 122 | 1.21 92 | 10.5 99 | 44.1 101 | 0.52 80 | 23.1 119 | 55.4 124 | 0.52 87 |
Adaptive [20] | 77.4 | 5.50 76 | 16.7 73 | 0.30 105 | 11.8 103 | 22.2 102 | 1.02 96 | 6.58 65 | 16.5 80 | 0.53 73 | 18.6 48 | 28.0 41 | 5.60 78 | 43.5 90 | 58.3 79 | 5.21 29 | 11.0 67 | 41.3 70 | 1.09 68 | 10.4 91 | 42.8 82 | 0.46 33 | 22.2 98 | 53.5 103 | 0.54 107 |
BriefMatch [124] | 79.0 | 5.45 68 | 16.5 62 | 0.31 109 | 9.84 59 | 19.6 52 | 1.43 118 | 7.55 110 | 15.6 52 | 1.08 115 | 20.3 104 | 29.2 89 | 7.97 127 | 43.3 50 | 58.3 79 | 5.43 95 | 12.0 117 | 41.5 79 | 2.37 129 | 9.84 31 | 41.5 32 | 0.56 93 | 21.4 19 | 51.7 20 | 0.52 87 |
Nguyen [33] | 79.1 | 5.63 94 | 15.9 34 | 0.23 80 | 13.8 119 | 23.8 117 | 1.37 114 | 6.89 86 | 18.7 113 | 0.59 84 | 20.8 111 | 30.8 108 | 5.44 64 | 43.1 25 | 58.1 46 | 5.14 16 | 10.6 19 | 40.4 40 | 0.93 24 | 11.9 127 | 45.9 116 | 0.73 115 | 22.0 81 | 52.8 78 | 0.52 87 |
2D-CLG [1] | 81.7 | 5.27 37 | 15.7 26 | 0.21 62 | 13.1 115 | 22.8 109 | 1.37 114 | 7.29 101 | 17.3 98 | 0.94 108 | 20.3 104 | 30.2 102 | 5.34 40 | 43.5 90 | 58.4 91 | 5.37 71 | 10.8 42 | 40.7 51 | 1.22 95 | 10.5 99 | 44.3 104 | 0.59 101 | 22.0 81 | 52.3 56 | 0.50 63 |
TV-L1-improved [17] | 82.2 | 5.26 35 | 16.0 41 | 0.28 98 | 11.6 98 | 22.0 98 | 1.06 101 | 7.21 98 | 16.3 70 | 0.79 102 | 18.8 65 | 28.5 61 | 5.70 86 | 43.5 90 | 58.5 99 | 5.22 31 | 11.0 67 | 41.5 79 | 1.05 59 | 10.4 91 | 44.6 107 | 0.74 117 | 22.1 89 | 53.2 94 | 0.53 96 |
SegOF [10] | 82.4 | 5.25 32 | 15.9 34 | 0.20 58 | 10.9 82 | 20.8 76 | 0.82 73 | 8.07 118 | 18.4 109 | 1.18 119 | 20.0 99 | 32.3 118 | 5.52 74 | 43.3 50 | 58.2 57 | 5.35 66 | 11.4 99 | 43.1 113 | 1.38 111 | 10.7 103 | 46.3 117 | 0.96 125 | 21.5 28 | 51.7 20 | 0.53 96 |
SPSA-learn [13] | 82.4 | 5.45 68 | 15.4 17 | 0.25 90 | 11.6 98 | 21.4 87 | 1.15 105 | 7.65 112 | 16.6 81 | 1.26 120 | 20.1 101 | 28.2 50 | 5.30 29 | 43.3 50 | 58.2 57 | 5.42 93 | 10.9 53 | 41.0 62 | 1.14 77 | 11.6 122 | 50.4 131 | 1.71 131 | 22.2 98 | 53.3 102 | 0.49 44 |
TriangleFlow [30] | 84.5 | 5.85 107 | 18.2 111 | 0.26 94 | 11.0 85 | 21.8 94 | 0.79 71 | 7.17 96 | 16.3 70 | 0.58 82 | 19.6 93 | 30.7 106 | 5.74 92 | 42.8 4 | 57.8 12 | 4.95 3 | 11.6 105 | 42.8 105 | 1.05 59 | 10.8 105 | 45.8 114 | 0.73 115 | 22.8 113 | 54.3 117 | 0.51 75 |
Black & Anandan [4] | 85.2 | 5.71 100 | 15.5 20 | 0.35 115 | 12.7 111 | 22.3 106 | 1.12 103 | 7.89 114 | 18.1 106 | 1.06 113 | 20.5 109 | 30.3 103 | 5.42 59 | 43.6 104 | 58.6 105 | 5.35 66 | 10.6 19 | 39.7 16 | 0.91 21 | 10.9 108 | 44.1 101 | 0.50 69 | 22.2 98 | 52.9 84 | 0.53 96 |
Rannacher [23] | 86.1 | 5.39 59 | 16.6 67 | 0.30 105 | 11.6 98 | 22.2 102 | 1.01 94 | 7.17 96 | 16.9 90 | 0.92 107 | 18.6 48 | 28.4 57 | 5.74 92 | 43.6 104 | 58.5 99 | 5.33 58 | 11.0 67 | 41.6 85 | 1.11 71 | 10.4 91 | 44.3 104 | 0.72 114 | 21.9 73 | 52.8 78 | 0.54 107 |
ROF-ND [107] | 86.4 | 6.15 113 | 16.4 59 | 0.14 8 | 10.4 74 | 21.1 79 | 0.70 63 | 7.09 92 | 15.9 58 | 0.40 47 | 20.7 110 | 32.9 121 | 5.82 98 | 43.4 67 | 58.2 57 | 5.37 71 | 11.6 105 | 43.4 119 | 1.16 82 | 11.6 122 | 46.4 118 | 0.55 90 | 22.6 108 | 53.8 106 | 0.54 107 |
Ad-TV-NDC [36] | 87.6 | 6.08 112 | 15.9 34 | 0.60 125 | 13.0 113 | 22.8 109 | 1.36 113 | 6.55 59 | 16.4 75 | 0.56 78 | 20.9 112 | 30.6 105 | 6.29 107 | 44.1 115 | 59.0 113 | 5.43 95 | 10.7 26 | 39.4 13 | 1.11 71 | 10.4 91 | 43.3 92 | 0.51 76 | 22.1 89 | 52.9 84 | 0.53 96 |
IAOF2 [51] | 89.5 | 6.17 114 | 18.3 112 | 0.30 105 | 12.0 105 | 23.3 114 | 0.93 86 | 5.90 5 | 16.1 65 | 0.42 53 | 20.4 108 | 31.2 115 | 5.75 94 | 43.7 109 | 58.9 111 | 5.39 80 | 11.2 87 | 42.0 93 | 1.08 64 | 10.3 82 | 42.7 80 | 0.48 56 | 22.7 110 | 54.2 114 | 0.52 87 |
Correlation Flow [75] | 90.1 | 5.61 90 | 17.8 106 | 0.15 15 | 10.8 79 | 21.7 91 | 0.82 73 | 6.40 46 | 14.8 30 | 0.42 53 | 19.1 85 | 29.0 80 | 6.04 105 | 43.9 112 | 58.6 105 | 6.05 127 | 12.0 117 | 43.9 123 | 1.29 102 | 11.0 110 | 45.3 111 | 0.70 112 | 22.5 104 | 54.1 112 | 0.51 75 |
Filter Flow [19] | 90.8 | 5.64 95 | 16.4 59 | 0.32 110 | 12.2 108 | 22.2 102 | 1.08 102 | 6.61 69 | 16.2 67 | 0.57 81 | 20.3 104 | 29.0 80 | 6.32 108 | 44.1 115 | 59.1 114 | 5.74 121 | 10.9 53 | 40.7 51 | 1.04 54 | 10.2 75 | 43.2 91 | 0.54 87 | 22.7 110 | 54.3 117 | 0.54 107 |
Bartels [41] | 92.5 | 5.52 78 | 17.2 87 | 0.40 121 | 10.0 62 | 20.7 73 | 0.94 87 | 6.50 54 | 15.8 56 | 0.54 74 | 19.9 95 | 30.0 100 | 7.79 124 | 44.8 125 | 59.2 115 | 6.72 130 | 12.8 129 | 42.4 98 | 3.06 131 | 10.0 55 | 42.0 49 | 0.54 87 | 22.1 89 | 53.2 94 | 0.54 107 |
Dynamic MRF [7] | 93.5 | 5.39 59 | 17.4 97 | 0.20 58 | 10.5 75 | 21.8 94 | 0.74 66 | 7.60 111 | 20.3 123 | 0.99 109 | 21.3 115 | 31.1 113 | 7.06 117 | 43.0 14 | 58.1 46 | 5.34 63 | 11.6 105 | 43.0 111 | 1.49 118 | 10.7 103 | 45.8 114 | 0.85 120 | 22.5 104 | 53.2 94 | 0.55 114 |
LocallyOriented [52] | 93.5 | 5.79 104 | 17.9 107 | 0.26 94 | 12.1 107 | 23.2 113 | 1.01 94 | 7.05 91 | 17.6 102 | 0.51 69 | 19.9 95 | 30.9 111 | 5.72 89 | 43.3 50 | 58.2 57 | 5.23 34 | 11.9 114 | 42.6 101 | 1.52 120 | 10.8 105 | 44.0 100 | 0.53 85 | 22.5 104 | 54.0 110 | 0.52 87 |
StereoOF-V1MT [119] | 96.2 | 5.94 109 | 18.8 115 | 0.20 58 | 11.3 91 | 22.6 108 | 0.94 87 | 7.95 115 | 19.6 117 | 1.00 110 | 21.6 116 | 30.7 106 | 6.76 113 | 43.3 50 | 58.3 79 | 5.37 71 | 12.1 120 | 42.6 101 | 1.82 126 | 11.6 122 | 46.7 120 | 0.90 122 | 21.8 62 | 51.8 28 | 0.50 63 |
ACK-Prior [27] | 97.1 | 5.46 70 | 17.7 104 | 0.15 15 | 9.70 52 | 20.3 69 | 0.67 59 | 7.76 113 | 16.4 75 | 1.08 115 | 19.9 95 | 31.0 112 | 6.01 104 | 44.7 124 | 59.6 121 | 5.78 122 | 12.1 120 | 44.2 127 | 1.33 105 | 10.6 101 | 44.2 103 | 0.53 85 | 23.4 124 | 56.1 128 | 0.52 87 |
TI-DOFE [24] | 98.4 | 6.39 116 | 18.7 114 | 0.36 118 | 14.8 126 | 25.5 128 | 1.66 124 | 7.45 108 | 20.2 121 | 0.78 101 | 22.8 122 | 32.5 119 | 6.04 105 | 43.2 40 | 58.4 91 | 5.17 20 | 10.9 53 | 40.4 40 | 0.92 22 | 11.2 115 | 45.6 113 | 0.65 106 | 23.2 121 | 54.2 114 | 0.65 125 |
UnFlow [129] | 99.1 | 6.39 116 | 20.9 119 | 0.21 62 | 13.0 113 | 24.4 124 | 1.15 105 | 8.06 117 | 21.1 124 | 0.82 105 | 19.2 87 | 29.6 95 | 5.64 81 | 43.1 25 | 58.0 30 | 5.40 87 | 11.8 113 | 42.8 105 | 1.36 109 | 11.0 110 | 42.4 61 | 0.70 112 | 24.3 130 | 54.8 121 | 0.70 127 |
StereoFlow [44] | 99.4 | 10.4 131 | 27.1 131 | 0.35 115 | 16.3 130 | 28.4 131 | 1.03 98 | 6.55 59 | 16.8 87 | 0.50 67 | 18.8 65 | 28.2 50 | 5.38 49 | 45.7 130 | 62.1 130 | 5.58 114 | 13.6 130 | 50.3 131 | 1.28 101 | 10.0 55 | 42.4 61 | 0.49 62 | 23.0 115 | 55.5 125 | 0.56 118 |
Horn & Schunck [3] | 99.5 | 5.81 106 | 17.3 91 | 0.21 62 | 13.1 115 | 23.5 115 | 1.26 109 | 8.03 116 | 19.7 118 | 1.08 115 | 22.6 120 | 32.7 120 | 5.59 77 | 43.6 104 | 58.7 108 | 5.39 80 | 10.9 53 | 40.6 48 | 1.02 48 | 11.7 125 | 46.5 119 | 0.60 102 | 22.8 113 | 53.9 109 | 0.55 114 |
2bit-BM-tele [98] | 99.5 | 5.61 90 | 15.9 34 | 0.50 124 | 11.5 96 | 21.9 96 | 1.04 99 | 6.57 62 | 15.1 41 | 0.79 102 | 20.1 101 | 29.8 97 | 7.50 121 | 44.8 125 | 59.6 121 | 6.26 128 | 12.2 123 | 42.8 105 | 2.11 128 | 11.2 115 | 49.2 130 | 1.26 129 | 21.8 62 | 52.1 44 | 0.55 114 |
NL-TV-NCC [25] | 111.2 | 6.44 118 | 20.3 118 | 0.24 86 | 10.7 77 | 22.1 100 | 0.68 61 | 7.38 106 | 17.2 96 | 0.59 84 | 22.2 119 | 34.7 126 | 6.82 114 | 45.5 129 | 60.2 128 | 6.68 129 | 12.3 126 | 44.6 129 | 1.19 87 | 14.4 131 | 48.1 127 | 0.67 107 | 24.0 129 | 56.4 129 | 0.55 114 |
SILK [79] | 111.9 | 6.21 115 | 19.3 117 | 0.39 120 | 13.8 119 | 24.0 119 | 1.73 125 | 8.85 123 | 20.2 121 | 1.41 122 | 21.8 117 | 31.1 113 | 7.10 118 | 43.5 90 | 58.5 99 | 5.45 99 | 11.9 114 | 41.4 74 | 2.03 127 | 10.8 105 | 45.5 112 | 0.77 118 | 22.4 103 | 53.2 94 | 0.60 121 |
HCIC-L [99] | 111.9 | 8.84 130 | 25.2 129 | 1.06 130 | 14.0 123 | 24.1 121 | 1.43 118 | 9.42 126 | 19.3 116 | 0.69 97 | 24.3 125 | 34.1 124 | 6.48 111 | 45.1 128 | 60.1 126 | 5.86 124 | 12.1 120 | 44.1 126 | 1.06 61 | 10.2 75 | 42.6 72 | 0.51 76 | 23.6 126 | 56.0 127 | 0.51 75 |
Adaptive flow [45] | 112.2 | 7.18 125 | 19.2 116 | 0.69 126 | 15.0 127 | 25.0 126 | 2.11 128 | 7.29 101 | 16.7 85 | 0.87 106 | 22.6 120 | 31.3 116 | 7.85 126 | 44.8 125 | 60.2 128 | 5.63 116 | 11.7 112 | 43.4 119 | 1.36 109 | 10.4 91 | 43.7 98 | 0.57 96 | 23.0 115 | 54.7 120 | 0.50 63 |
Learning Flow [11] | 115.0 | 5.91 108 | 18.6 113 | 0.30 105 | 12.0 105 | 22.9 111 | 1.00 92 | 8.30 121 | 20.0 120 | 1.33 121 | 21.9 118 | 32.9 121 | 6.94 116 | 44.5 123 | 59.7 124 | 5.97 126 | 11.5 102 | 42.6 101 | 1.35 107 | 11.3 117 | 46.8 121 | 0.69 111 | 23.7 127 | 55.9 126 | 0.62 123 |
GroupFlow [9] | 115.2 | 7.04 123 | 22.5 125 | 0.28 98 | 12.5 110 | 24.0 119 | 1.13 104 | 9.10 124 | 22.0 126 | 1.45 123 | 21.0 114 | 33.6 123 | 5.93 102 | 44.1 115 | 59.3 119 | 5.50 107 | 12.2 123 | 44.4 128 | 1.42 114 | 11.1 114 | 45.2 110 | 0.61 105 | 22.7 110 | 54.1 112 | 0.56 118 |
SLK [47] | 116.6 | 6.55 119 | 21.1 121 | 0.32 110 | 13.5 118 | 23.1 112 | 1.44 120 | 9.16 125 | 21.2 125 | 1.49 127 | 24.9 127 | 34.2 125 | 7.81 125 | 43.5 90 | 58.8 109 | 5.34 63 | 12.2 123 | 43.1 113 | 1.45 116 | 11.9 127 | 48.9 129 | 0.96 125 | 23.0 115 | 54.0 110 | 0.64 124 |
Heeger++ [104] | 118.0 | 7.79 128 | 25.2 129 | 0.17 35 | 13.9 122 | 24.2 122 | 1.33 111 | 11.8 129 | 28.7 130 | 1.49 127 | 23.4 123 | 30.8 108 | 7.63 122 | 44.4 122 | 59.9 125 | 5.62 115 | 12.6 128 | 43.1 113 | 1.77 125 | 12.6 130 | 46.9 122 | 0.87 121 | 23.2 121 | 53.5 103 | 0.60 121 |
FFV1MT [106] | 118.6 | 6.93 122 | 22.8 126 | 0.24 86 | 14.0 123 | 23.5 115 | 1.48 122 | 11.2 128 | 27.7 129 | 1.52 129 | 23.4 123 | 30.8 108 | 7.63 122 | 44.0 113 | 59.2 115 | 5.69 118 | 12.0 117 | 41.6 85 | 1.56 121 | 12.1 129 | 47.3 125 | 0.95 124 | 23.4 124 | 54.2 114 | 0.79 129 |
FOLKI [16] | 121.6 | 7.10 124 | 21.1 121 | 0.94 129 | 15.3 128 | 25.5 128 | 2.28 129 | 8.49 122 | 22.2 127 | 1.47 126 | 26.3 129 | 35.2 128 | 10.6 130 | 44.0 113 | 59.6 121 | 5.54 111 | 11.6 105 | 41.8 91 | 1.49 118 | 11.4 119 | 47.7 126 | 0.90 122 | 23.3 123 | 54.9 122 | 0.67 126 |
Pyramid LK [2] | 122.9 | 7.19 126 | 21.0 120 | 0.93 128 | 16.2 129 | 25.1 127 | 2.91 130 | 14.0 130 | 18.5 111 | 2.57 130 | 32.5 131 | 46.2 131 | 13.7 131 | 44.2 119 | 60.1 126 | 5.48 103 | 11.6 105 | 42.5 100 | 1.40 113 | 11.4 119 | 47.2 124 | 1.28 130 | 23.7 127 | 56.7 130 | 1.08 130 |
PGAM+LK [55] | 123.2 | 7.51 127 | 23.5 128 | 0.73 127 | 13.8 119 | 24.2 122 | 1.92 127 | 9.44 127 | 22.7 128 | 1.45 123 | 26.4 130 | 36.9 129 | 10.5 129 | 44.1 115 | 59.5 120 | 5.72 120 | 12.4 127 | 44.0 124 | 1.75 124 | 11.3 117 | 47.0 123 | 0.68 110 | 23.0 115 | 54.5 119 | 0.76 128 |
Periodicity [78] | 129.5 | 8.05 129 | 23.2 127 | 1.34 131 | 20.5 131 | 27.4 130 | 3.39 131 | 15.2 131 | 30.5 131 | 4.22 131 | 26.2 128 | 43.5 130 | 9.47 128 | 46.4 131 | 62.7 131 | 6.92 131 | 13.7 131 | 44.6 129 | 2.88 130 | 11.4 119 | 48.3 128 | 1.18 128 | 25.7 131 | 59.2 131 | 1.29 131 |
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. |