Optical flow evaluation results Statistics:     Average   SD   R2.5   R5.0   R10.0   A50   A75   A95  
Error type:   endpoint   angle   interpolation   normalized interpolation  
Show images: below table   above table        
Average
angle
error
avg. Army
(Hidden texture)
GT   im0   im1
Mequon
(Hidden texture)
GT   im0   im1
Schefflera
(Hidden texture)
GT   im0   im1
Wooden
(Hidden texture)
GT   im0   im1
Grove
(Synthetic)
GT   im0   im1
Urban
(Synthetic)
GT   im0   im1
Yosemite
(Synthetic)
GT   im0   im1
Teddy
(Stereo)
GT   im0   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
NNF-Local [87]5.0 2.69 3 7.56 4 1.98 3 1.97 4 7.01 4 1.59 4 2.18 2 5.36 3 1.53 4 1.87 2 9.14 5 1.06 4 2.28 2 2.94 1 1.57 2 2.39 5 6.78 2 2.15 9 2.00 20 3.36 17 1.62 14 0.99 1 2.16 2 0.57 2
NN-field [71]9.8 2.89 8 8.13 16 2.11 5 2.10 6 7.15 9 1.77 14 2.27 4 5.59 5 1.61 8 1.58 1 8.52 4 0.79 1 2.35 4 3.05 5 1.60 3 1.89 1 5.20 1 1.37 1 2.43 46 3.70 49 1.95 37 1.01 2 2.25 3 0.53 1
OFLAF [77]12.5 3.04 15 7.80 10 2.40 13 2.14 7 7.02 5 1.72 9 2.25 3 5.32 2 1.56 5 2.62 17 13.7 23 1.37 19 2.35 4 3.13 6 1.62 4 2.98 22 7.73 7 2.57 20 2.08 25 3.27 11 2.05 40 1.33 12 2.43 6 1.40 15
PMMST [114]13.5 3.42 41 7.60 5 2.65 27 2.32 11 6.39 1 2.20 31 2.63 11 6.08 8 2.03 25 2.06 4 6.07 1 1.44 26 2.60 10 3.27 8 1.91 10 2.56 7 6.78 2 2.09 5 2.06 22 3.53 36 1.63 15 1.27 9 2.29 4 1.02 6
nLayers [57]16.0 2.80 6 7.42 3 2.20 8 2.71 29 7.24 10 2.55 59 2.61 9 6.24 9 2.45 49 2.30 10 12.7 11 1.16 7 2.30 3 3.02 3 1.70 5 2.62 10 6.95 4 2.09 5 2.29 40 3.46 26 1.89 34 1.38 14 3.06 17 1.29 13
MDP-Flow2 [68]18.1 3.23 30 7.93 13 2.60 19 1.92 2 6.64 2 1.52 1 2.46 7 5.91 7 1.56 5 3.05 41 15.8 50 1.51 36 2.77 23 3.50 17 2.16 25 2.86 18 8.58 18 2.70 31 2.00 20 3.50 33 1.59 12 1.28 10 2.67 11 0.89 4
ComponentFusion [96]18.5 2.78 5 8.20 17 2.05 4 2.04 5 7.31 11 1.66 8 2.55 8 6.78 13 1.61 8 2.24 9 13.1 13 1.01 3 2.71 20 3.56 19 2.10 21 3.55 50 12.4 54 3.22 56 2.19 35 3.60 42 1.54 11 1.32 11 2.91 13 1.13 8
TC/T-Flow [76]20.8 2.69 3 7.75 9 1.87 2 2.76 32 10.2 45 1.73 10 3.33 24 9.01 30 1.49 2 2.86 32 16.7 60 1.21 9 2.60 10 3.49 16 1.90 9 2.21 2 7.65 5 2.04 4 1.84 10 3.23 8 3.14 85 2.03 37 4.53 37 1.49 19
FC-2Layers-FF [74]24.0 3.02 14 7.87 12 2.61 20 2.72 30 9.35 35 2.29 39 2.36 5 5.47 4 2.15 32 2.48 11 12.6 10 1.28 11 2.49 7 3.19 7 2.03 16 3.39 39 8.92 20 2.83 41 2.83 69 3.92 62 2.80 64 1.25 7 2.57 10 1.20 10
WLIF-Flow [93]25.1 2.96 10 7.67 6 2.40 13 2.41 16 7.70 15 2.10 25 2.98 17 7.63 18 1.97 24 2.71 24 13.5 19 1.33 14 3.01 40 4.00 46 2.40 42 3.03 25 8.32 12 2.44 16 2.09 27 3.36 17 2.04 39 2.26 44 4.97 44 2.59 49
Layers++ [37]25.5 3.11 17 8.22 20 2.79 39 2.43 19 7.02 5 2.24 34 2.43 6 5.77 6 2.18 35 2.13 6 9.71 7 1.15 6 2.35 4 3.02 3 1.96 11 3.81 56 11.4 41 3.22 56 2.74 64 4.01 67 2.35 50 1.45 15 3.05 16 1.79 28
HAST [109]25.6 2.58 1 7.12 1 1.81 1 2.41 16 7.05 7 2.10 25 1.83 1 4.19 1 1.17 1 2.84 31 15.5 45 1.08 5 2.23 1 2.97 2 1.40 1 3.72 55 10.0 31 3.92 77 3.40 88 4.90 93 5.66 114 1.20 6 2.09 1 1.24 11
FESL [72]27.2 2.96 10 7.70 7 2.54 16 3.26 67 10.4 46 2.56 60 3.25 22 8.39 22 2.17 33 2.56 13 13.2 14 1.31 13 2.57 9 3.40 11 2.12 24 2.60 9 7.65 5 2.30 10 2.64 60 4.22 75 2.47 52 1.75 26 3.49 25 1.71 24
AGIF+OF [85]27.3 3.06 16 8.20 17 2.55 18 3.17 56 10.6 49 2.46 52 3.46 29 8.97 29 2.24 38 2.61 15 13.7 23 1.33 14 2.63 14 3.46 14 2.11 22 2.88 20 8.34 14 2.35 12 2.10 29 3.56 38 2.09 42 1.80 28 3.68 28 2.24 38
Efficient-NL [60]27.4 2.99 13 8.23 21 2.28 9 2.72 30 8.95 31 2.25 37 3.81 38 9.87 36 2.07 29 2.77 28 14.3 30 1.46 31 2.61 12 3.48 15 1.96 11 3.31 35 8.33 13 2.59 22 2.60 55 3.75 50 2.54 55 1.60 21 3.02 14 1.66 21
LME [70]27.5 3.15 22 8.04 15 2.31 11 1.95 3 6.65 3 1.59 4 4.03 44 9.31 31 4.57 90 2.69 22 13.6 21 1.42 23 2.85 30 3.61 22 2.42 43 3.47 46 12.8 59 3.17 52 2.12 31 3.53 36 1.73 17 1.34 13 2.75 12 1.18 9
ALD-Flow [66]27.6 2.82 7 7.86 11 2.16 6 2.84 38 10.1 42 1.86 16 3.73 36 10.4 39 1.67 11 3.10 43 16.8 61 1.28 11 2.69 19 3.60 21 1.85 8 2.79 14 11.3 40 2.32 11 2.07 24 3.25 10 3.10 82 2.03 37 5.11 45 1.94 31
RNLOD-Flow [121]28.6 2.66 2 7.33 2 2.17 7 2.53 25 9.46 36 1.86 16 3.94 42 10.7 45 1.95 22 2.50 12 13.5 19 1.21 9 2.68 17 3.62 24 2.05 18 2.99 23 8.59 19 2.75 35 3.00 77 4.54 82 3.25 90 1.48 17 3.24 20 1.76 27
IROF++ [58]29.0 3.17 24 8.69 28 2.61 20 2.79 34 9.61 37 2.33 40 3.43 26 8.86 26 2.38 44 2.87 33 14.8 35 1.52 38 2.74 21 3.57 20 2.19 26 3.20 32 9.70 28 2.71 32 1.96 18 3.45 25 1.22 5 1.80 28 4.06 30 2.50 45
NNF-EAC [103]29.1 3.31 33 8.21 19 2.68 29 2.19 9 7.49 13 1.76 12 2.73 13 6.62 12 1.70 12 3.18 48 15.8 50 1.64 47 2.87 32 3.66 27 2.24 28 3.02 24 8.07 10 2.59 22 2.19 35 3.48 29 1.74 18 2.85 58 6.52 57 3.12 61
PH-Flow [101]29.5 3.19 27 8.87 33 2.71 30 2.84 38 9.33 34 2.37 42 2.85 14 7.20 15 2.36 41 2.92 36 15.4 42 1.51 36 2.63 14 3.42 12 2.04 17 3.03 25 8.52 17 2.49 18 2.69 62 3.60 42 3.13 84 1.25 7 2.53 8 1.34 14
Classic+CPF [83]30.0 3.14 20 8.60 26 2.63 24 3.03 54 10.6 49 2.33 40 3.66 33 9.58 32 2.20 36 2.61 15 14.1 28 1.34 17 2.68 17 3.53 18 2.21 27 2.85 17 7.95 9 2.38 13 2.44 48 3.49 31 2.90 75 1.67 24 3.40 23 2.43 44
Sparse-NonSparse [56]31.5 3.14 20 8.75 30 2.76 37 3.02 52 10.6 49 2.43 47 3.45 28 8.96 27 2.36 41 2.66 19 13.7 23 1.42 23 2.85 30 3.75 33 2.33 33 3.28 34 9.40 25 2.73 33 2.42 45 3.31 13 2.69 59 1.47 16 3.07 18 1.66 21
TC-Flow [46]32.8 2.91 9 8.00 14 2.34 12 2.18 8 8.77 26 1.52 1 3.84 40 10.7 45 1.49 2 3.13 44 16.6 59 1.46 31 2.78 24 3.73 32 1.96 11 3.08 28 11.4 41 2.66 26 1.94 16 3.43 22 3.20 89 3.06 62 7.04 60 4.08 86
LSM [39]33.8 3.12 18 8.62 27 2.75 36 3.00 50 10.5 48 2.44 49 3.43 26 8.85 25 2.35 40 2.66 19 13.6 21 1.44 26 2.82 26 3.68 28 2.36 35 3.38 38 9.41 26 2.81 39 2.69 62 3.52 34 2.84 68 1.59 20 3.38 22 1.80 29
SVFilterOh [111]34.3 3.63 47 8.82 31 2.86 41 2.60 27 8.06 18 2.05 24 2.95 15 7.09 14 2.03 25 2.80 30 13.8 26 1.41 22 2.63 14 3.42 12 1.75 7 3.49 47 10.3 33 3.23 58 3.63 96 5.75 114 4.47 107 1.09 4 2.45 7 0.92 5
Correlation Flow [75]34.4 3.38 39 8.40 23 2.64 25 2.23 10 7.54 14 1.56 3 5.14 65 13.1 64 1.60 7 2.09 5 8.15 3 1.35 18 3.12 47 4.09 52 2.34 34 4.01 66 11.5 45 4.00 78 2.59 54 3.61 44 3.00 80 1.49 18 3.04 15 1.42 17
Ramp [62]34.8 3.18 26 8.83 32 2.73 33 2.89 43 10.1 42 2.44 49 3.27 23 8.43 23 2.38 44 2.74 26 14.2 29 1.46 31 2.82 26 3.69 30 2.29 31 3.37 37 9.31 23 2.93 45 2.62 58 3.38 20 3.19 88 1.54 19 3.21 19 2.24 38
PMF [73]35.2 3.61 45 9.07 36 2.62 22 2.40 14 8.05 17 1.83 15 2.61 9 6.27 10 1.65 10 3.35 57 15.4 42 1.58 42 2.54 8 3.27 8 1.71 6 3.59 51 11.1 39 3.46 64 4.07 105 6.18 119 4.02 105 1.06 3 2.38 5 1.25 12
ProbFlowFields [128]35.5 4.18 65 12.4 78 3.40 72 2.43 19 8.16 20 2.19 30 3.65 32 9.72 34 2.86 64 2.22 7 9.42 6 1.42 23 3.01 40 3.96 43 2.36 35 2.73 13 10.9 34 2.51 19 1.89 15 3.39 21 1.82 23 2.59 51 6.21 55 2.75 52
COFM [59]35.6 3.17 24 9.90 51 2.46 15 2.41 16 8.34 23 1.92 19 3.77 37 10.5 40 2.54 52 2.71 24 14.9 37 1.19 8 3.08 45 3.92 41 3.25 83 3.83 58 10.9 34 3.15 51 2.20 38 3.35 15 2.91 77 1.62 23 2.56 9 2.09 35
FMOF [94]36.8 3.12 18 8.23 21 2.73 33 3.25 64 10.7 56 2.52 57 3.01 18 7.61 17 2.20 36 2.56 13 13.4 17 1.33 14 2.75 22 3.61 22 2.24 28 3.66 53 8.50 16 2.78 37 2.62 58 3.84 57 3.27 92 2.66 54 5.69 48 1.95 33
OAR-Flow [125]37.9 3.37 37 9.87 50 2.67 28 4.22 85 12.8 80 2.87 77 4.95 61 13.4 67 2.66 56 3.23 50 16.4 58 1.37 19 2.83 28 3.82 36 1.97 14 2.49 6 10.9 34 1.87 3 1.52 2 2.82 1 1.86 29 1.85 31 4.35 35 1.68 23
Classic+NL [31]38.0 3.20 29 8.72 29 2.81 40 3.02 52 10.6 49 2.44 49 3.46 29 8.84 24 2.38 44 2.78 29 14.3 30 1.46 31 2.83 28 3.68 28 2.31 32 3.40 40 9.09 22 2.76 36 2.87 71 3.82 56 2.86 72 1.67 24 3.53 26 2.26 41
TV-L1-MCT [64]39.1 3.16 23 8.48 25 2.71 30 3.28 68 10.8 60 2.60 67 3.95 43 10.5 40 2.38 44 2.69 22 13.9 27 1.45 30 2.94 36 3.79 34 2.63 61 3.50 48 9.75 29 3.06 49 2.08 25 3.35 15 2.29 48 1.95 34 3.89 29 2.71 51
CostFilter [40]42.8 3.84 50 9.64 46 3.06 50 2.55 26 8.09 19 2.03 22 2.69 12 6.47 11 1.88 18 3.66 68 16.8 61 1.88 59 2.62 13 3.34 10 1.99 15 4.05 67 11.0 38 3.65 71 4.16 107 7.18 126 4.66 109 1.16 5 3.36 21 0.87 3
SimpleFlow [49]43.0 3.35 34 9.20 39 2.98 48 3.18 59 10.7 56 2.71 70 5.06 63 12.6 62 2.70 58 2.95 38 15.1 39 1.58 42 2.91 35 3.79 34 2.47 45 3.59 51 9.49 27 2.99 47 2.39 43 3.46 26 2.24 47 1.60 21 3.56 27 1.57 20
2DHMM-SAS [92]44.9 3.19 27 8.89 34 2.71 30 3.20 61 11.5 67 2.38 43 5.19 66 12.2 57 2.73 60 2.92 36 15.2 40 1.53 39 2.79 25 3.65 26 2.27 30 3.45 44 9.34 24 2.78 37 2.66 61 3.56 38 3.07 81 2.34 47 5.12 46 2.97 59
MLDP_OF [89]45.9 4.13 61 10.3 58 3.60 80 2.34 12 7.70 15 1.88 18 4.23 49 10.9 48 1.87 17 2.74 26 14.6 34 1.37 19 3.10 46 3.91 40 2.48 49 3.40 40 9.00 21 3.79 74 3.46 90 4.20 73 5.55 113 2.31 45 4.64 40 1.98 34
S2D-Matching [84]46.1 3.36 35 9.66 47 2.86 41 3.19 60 11.1 63 2.46 52 4.86 60 12.9 63 2.47 50 2.67 21 13.2 14 1.44 26 2.87 32 3.72 31 2.38 38 3.45 44 9.76 30 2.95 46 3.05 78 3.79 54 3.30 94 1.95 34 4.16 33 3.00 60
MDP-Flow [26]46.2 3.48 43 9.46 43 3.10 52 2.45 21 7.36 12 2.41 44 3.21 21 8.31 21 2.78 62 3.18 48 17.8 66 1.70 52 3.03 42 3.87 37 2.60 57 3.43 42 12.6 57 2.81 39 2.19 35 3.88 60 1.60 13 4.13 78 9.96 83 3.86 82
FlowFields+ [130]46.3 4.57 83 13.7 92 3.35 64 2.94 48 10.1 42 2.58 64 4.05 45 10.6 42 3.26 73 2.90 35 13.2 14 1.81 57 3.18 51 4.20 59 2.54 51 2.68 12 11.4 41 2.40 15 1.84 10 3.62 45 1.77 19 2.48 48 5.86 49 2.77 53
AggregFlow [97]47.0 4.25 71 11.9 76 3.26 56 4.46 90 13.7 89 3.43 87 4.76 58 12.4 58 3.93 87 3.28 53 15.6 46 1.68 49 2.89 34 3.89 38 2.08 19 2.32 3 7.75 8 2.14 7 2.06 22 3.77 52 1.48 10 2.07 41 4.11 31 2.36 42
CombBMOF [113]47.2 3.94 54 10.6 62 2.74 35 2.80 35 8.55 25 2.16 28 3.10 20 7.99 20 1.76 13 2.99 39 13.4 17 1.95 64 3.04 43 3.89 38 2.49 50 5.64 94 12.3 52 6.74 108 3.54 92 5.16 101 2.81 65 1.85 31 4.60 39 1.10 7
IROF-TV [53]47.2 3.40 40 9.29 41 2.95 47 2.99 49 11.1 63 2.53 58 3.81 38 9.81 35 2.44 48 3.25 52 16.9 63 1.78 56 3.27 65 4.10 53 2.93 75 4.47 74 16.0 86 3.53 66 1.70 4 3.21 6 1.12 3 1.91 33 4.75 42 2.19 37
S2F-IF [123]48.8 4.51 81 13.6 90 3.31 60 2.90 44 10.4 46 2.48 55 4.07 47 10.8 47 3.15 70 3.31 54 15.7 49 1.90 60 3.17 49 4.19 57 2.55 54 2.81 16 11.6 47 2.60 24 1.86 13 3.67 47 1.87 30 2.11 43 4.64 40 2.54 48
NL-TV-NCC [25]51.1 3.89 52 9.16 38 2.98 48 2.87 42 9.69 38 1.99 20 4.44 53 11.6 52 1.76 13 2.64 18 11.8 9 1.48 35 3.49 80 4.60 87 2.47 45 4.67 81 13.5 64 4.26 86 2.83 69 4.57 84 2.84 68 2.62 52 6.00 53 2.25 40
FlowFields [110]51.4 4.57 83 13.7 92 3.38 67 3.01 51 10.6 49 2.59 65 4.19 48 11.1 49 3.30 74 3.17 47 15.0 38 1.96 65 3.21 58 4.24 66 2.61 60 2.91 21 12.4 54 2.66 26 1.84 10 3.46 26 1.84 27 2.50 49 6.15 54 2.79 54
Sparse Occlusion [54]51.7 3.62 46 9.12 37 2.90 43 2.92 46 9.08 32 2.56 60 4.49 55 11.8 55 2.11 31 3.14 45 15.8 50 1.57 41 3.26 63 4.22 62 2.36 35 3.52 49 10.9 34 2.66 26 5.10 122 6.32 120 3.15 86 2.02 36 4.92 43 1.71 24
EPPM w/o HM [88]52.0 4.25 71 11.1 66 3.13 53 2.36 13 8.35 24 1.76 12 3.72 35 10.2 38 1.81 15 3.24 51 14.5 33 1.94 62 3.16 48 3.94 42 2.82 70 4.78 84 12.9 60 4.32 87 3.64 98 4.54 82 5.73 115 1.76 27 4.11 31 1.94 31
OFH [38]52.2 3.90 53 9.77 49 3.62 83 2.84 38 11.0 62 2.04 23 5.52 71 14.4 72 1.89 19 3.52 60 20.5 83 1.60 45 3.18 51 4.06 51 2.82 70 3.86 59 14.1 72 3.59 68 1.77 6 3.62 45 1.81 22 2.64 53 7.08 62 2.15 36
PGM-C [120]53.2 4.62 88 14.0 97 3.39 69 3.29 70 12.3 72 2.70 69 4.39 52 11.7 53 3.43 77 4.00 76 19.8 74 2.15 70 3.19 53 4.23 63 2.54 51 2.79 14 11.9 49 2.45 17 1.83 8 3.21 6 1.83 25 2.31 45 5.87 50 1.82 30
Occlusion-TV-L1 [63]54.2 3.59 44 9.61 44 2.64 25 2.93 47 10.6 49 2.41 44 6.16 77 15.2 75 2.70 58 3.32 55 17.0 64 1.68 49 3.38 70 4.44 77 2.82 70 3.10 30 13.2 63 2.68 29 2.17 32 3.52 34 1.46 8 4.63 88 11.1 97 3.53 71
Complementary OF [21]55.1 4.44 77 11.2 69 4.04 90 2.51 24 9.77 40 1.74 11 3.93 41 10.6 42 2.04 27 3.87 72 18.8 69 2.19 73 3.17 49 4.00 46 2.92 74 4.64 79 13.8 69 3.64 70 2.17 32 3.36 17 2.51 53 3.08 63 7.04 60 3.65 75
Adaptive [20]56.1 3.29 31 9.43 42 2.28 9 3.10 55 11.4 66 2.46 52 6.58 81 15.7 81 2.52 51 3.14 45 15.6 46 1.56 40 3.67 89 4.46 79 3.48 92 3.32 36 13.0 62 2.38 13 2.76 67 4.39 79 1.93 36 3.58 70 8.18 69 2.88 56
Aniso-Texture [82]56.2 2.96 10 7.72 8 2.54 16 2.48 23 8.26 22 2.24 34 6.48 79 15.9 85 2.63 54 1.96 3 10.1 8 0.98 2 3.26 63 4.21 60 2.60 57 5.74 96 16.9 94 5.61 99 4.47 114 5.88 117 3.33 95 3.51 69 7.12 63 3.68 77
ACK-Prior [27]56.7 4.19 67 9.27 40 3.60 80 2.40 14 8.21 21 1.65 7 3.40 25 8.96 27 1.84 16 2.87 33 14.4 32 1.44 26 3.36 69 4.15 54 3.07 78 6.35 104 16.1 88 4.90 93 4.21 109 4.80 88 6.03 117 3.29 66 5.99 52 2.82 55
DPOF [18]58.9 4.67 92 12.6 83 3.30 58 3.57 76 10.6 49 3.12 84 3.09 19 7.50 16 2.32 39 3.06 42 14.8 35 1.82 58 3.21 58 4.18 56 2.79 69 4.47 74 12.5 56 3.33 59 4.09 106 3.92 62 6.96 119 2.09 42 4.39 36 1.74 26
CPM-Flow [116]58.9 4.63 89 14.1 100 3.39 69 3.33 71 12.5 76 2.73 71 4.37 50 11.7 53 3.43 77 4.00 76 19.9 76 2.14 69 3.19 53 4.23 63 2.54 51 3.08 28 12.0 50 2.88 43 1.87 14 3.44 23 1.84 27 2.91 60 7.48 67 2.91 58
EpicFlow [102]59.2 4.61 87 14.0 97 3.39 69 3.33 71 12.5 76 2.74 72 5.37 68 14.8 74 3.46 80 3.94 75 19.2 71 2.13 68 3.20 55 4.23 63 2.58 56 2.87 19 12.2 51 2.64 25 1.83 8 3.28 12 1.83 25 3.21 65 7.12 63 3.61 72
Kuang [131]59.6 4.36 74 13.6 90 3.21 55 3.21 62 12.5 76 2.51 56 4.46 54 12.4 58 3.07 67 3.54 61 17.8 66 1.94 62 3.29 66 4.34 70 2.69 65 4.16 71 14.2 73 4.09 80 1.77 6 3.34 14 1.82 23 2.73 56 6.78 58 3.40 68
DeepFlow2 [108]59.8 4.04 58 11.2 69 3.38 67 3.80 78 12.4 75 2.86 76 5.12 64 13.4 67 3.00 65 4.17 81 20.1 78 2.18 72 2.96 37 3.97 44 2.08 19 3.06 27 12.6 57 2.69 30 2.17 32 3.24 9 2.71 60 4.74 90 10.4 90 4.38 91
ROF-ND [107]60.3 4.12 59 10.0 52 3.37 66 2.78 33 8.82 28 2.12 27 4.61 57 11.9 56 2.09 30 2.23 8 6.56 2 1.69 51 3.60 86 4.75 97 2.85 73 4.92 87 13.6 67 3.75 72 4.59 116 5.18 102 4.10 106 2.67 55 5.19 47 3.46 70
TCOF [69]60.5 4.17 64 10.4 60 3.71 85 3.17 56 10.7 56 2.59 65 6.58 81 15.7 81 3.82 85 3.69 70 16.1 55 2.37 80 3.78 92 4.95 106 2.47 45 2.59 8 8.47 15 2.58 21 3.66 99 4.83 89 2.67 58 1.83 30 4.20 34 1.46 18
RFlow [90]61.5 3.82 49 10.0 52 3.44 75 2.61 28 9.73 39 2.02 21 5.66 73 14.5 73 2.05 28 3.93 74 23.1 97 1.90 60 3.24 60 4.19 57 2.66 64 4.12 70 15.2 82 3.34 61 2.61 56 3.56 38 2.65 57 4.48 84 10.5 93 3.93 85
Steered-L1 [118]62.1 3.30 32 8.44 24 2.91 44 1.89 1 7.14 8 1.60 6 3.61 31 9.91 37 1.89 19 3.45 58 19.4 73 1.64 47 3.42 73 4.30 68 3.39 86 5.18 89 14.5 75 4.37 89 5.09 121 5.05 97 10.1 123 5.56 97 10.2 88 6.24 104
HBM-GC [105]62.2 5.25 97 10.5 61 4.34 97 3.17 56 8.78 27 2.94 80 4.38 51 10.6 42 2.68 57 3.59 64 12.8 12 2.47 83 2.96 37 3.64 25 2.64 62 3.96 65 8.26 11 3.56 67 4.40 112 5.92 118 3.62 99 2.55 50 6.34 56 3.29 64
ComplOF-FED-GPU [35]64.2 4.28 73 11.3 71 3.70 84 3.25 64 13.0 82 2.16 28 4.06 46 11.2 50 1.95 22 3.91 73 19.2 71 2.01 66 3.20 55 4.15 54 2.64 62 4.61 78 16.1 88 3.90 76 2.98 76 3.77 52 3.69 100 2.85 58 7.44 66 2.53 47
SRR-TVOF-NL [91]64.4 4.47 79 10.9 64 3.32 62 4.04 82 13.2 85 2.90 78 4.81 59 12.5 60 3.15 70 3.33 56 15.3 41 1.61 46 3.24 60 4.03 50 2.70 67 3.94 63 11.8 48 3.33 59 4.16 107 5.21 105 3.44 98 2.06 40 3.48 24 2.42 43
TF+OM [100]66.5 3.97 55 10.2 55 2.94 46 2.91 45 9.12 33 2.57 63 5.22 67 11.5 51 6.92 95 3.59 64 16.1 55 2.28 77 3.20 55 3.97 44 3.11 79 4.70 82 14.5 75 4.32 87 3.06 80 4.84 90 2.71 60 3.93 74 8.79 74 4.32 90
Aniso. Huber-L1 [22]67.2 3.71 48 10.1 54 3.08 51 4.36 89 13.0 82 3.77 91 6.92 85 15.3 77 3.60 83 3.54 61 15.9 53 2.04 67 3.38 70 4.45 78 2.47 45 3.88 60 12.9 60 2.74 34 3.37 87 4.36 78 2.85 71 3.16 64 7.52 68 2.90 57
DeepFlow [86]68.5 4.49 80 11.7 73 4.14 92 4.26 86 12.8 80 3.36 85 5.96 74 14.2 71 5.10 91 4.89 92 23.1 97 2.67 86 2.98 39 4.00 46 2.11 22 3.26 33 13.5 64 2.84 42 2.09 27 3.10 3 2.77 62 5.83 99 11.4 99 5.45 101
TV-L1-improved [17]69.0 3.36 35 9.63 45 2.62 22 2.82 36 10.7 56 2.23 32 6.50 80 15.8 83 2.73 60 3.80 71 21.3 88 1.76 55 3.34 68 4.38 75 2.39 39 5.97 98 18.1 100 5.67 100 3.57 94 4.92 95 3.43 97 4.01 77 9.84 82 3.44 69
Classic++ [32]69.0 3.37 37 9.67 48 2.91 44 3.28 68 12.1 70 2.61 68 5.46 70 14.1 70 3.00 65 3.63 66 20.2 81 1.70 52 3.24 60 4.34 70 2.60 57 4.65 80 16.0 86 3.60 69 3.09 81 3.94 65 3.28 93 4.64 89 10.4 90 3.71 78
LocallyOriented [52]71.9 4.54 82 12.8 85 3.27 57 4.73 94 14.8 96 3.73 90 7.77 91 18.3 98 3.44 79 3.56 63 15.6 46 2.22 74 3.46 77 4.47 80 2.69 65 3.15 31 10.2 32 3.19 54 2.61 56 4.20 73 2.52 54 4.39 81 8.52 71 5.23 97
SIOF [67]72.1 4.23 69 10.2 55 3.31 60 3.97 80 14.5 94 2.97 81 7.81 93 16.4 88 7.48 96 4.82 88 20.1 78 2.96 89 3.54 83 4.49 81 3.12 80 4.31 72 13.5 64 4.13 82 2.36 42 3.59 41 1.68 16 3.46 68 7.39 65 3.37 66
TriangleFlow [30]73.9 4.12 59 10.6 62 3.47 76 3.47 75 13.1 84 2.41 44 6.00 75 15.2 75 2.17 33 2.99 39 16.0 54 1.58 42 4.46 113 5.79 118 4.15 105 5.42 92 13.9 71 5.24 94 3.10 83 5.47 110 2.90 75 3.02 61 6.82 59 3.64 74
CRTflow [80]74.5 4.18 65 11.8 75 3.20 54 3.22 63 10.8 60 2.43 47 6.20 78 15.5 79 2.63 54 4.21 82 22.0 91 2.24 75 3.32 67 4.34 70 2.44 44 7.43 111 19.3 106 8.15 114 2.55 52 4.09 69 2.59 56 4.60 86 11.2 98 4.45 92
Brox et al. [5]74.8 4.44 77 12.4 78 4.22 95 3.72 77 13.5 88 3.06 82 4.97 62 13.3 66 3.11 68 4.58 86 22.0 91 2.37 80 3.79 94 4.60 87 4.33 109 3.91 62 17.0 95 3.45 63 2.22 39 3.79 54 1.19 4 4.62 87 10.0 84 3.38 67
BriefMatch [124]75.9 3.44 42 9.01 35 2.77 38 2.85 41 9.93 41 2.23 32 2.97 16 7.65 19 1.94 21 3.64 67 20.1 78 1.75 54 4.10 107 4.90 104 5.82 119 7.95 113 17.8 98 8.08 113 4.73 118 5.20 103 12.2 125 7.88 115 12.0 102 13.7 121
Rannacher [23]76.8 4.13 61 11.0 65 3.61 82 3.39 73 12.3 72 2.80 75 7.26 87 17.4 94 3.59 82 4.40 84 23.1 97 2.24 75 3.43 75 4.54 84 2.56 55 5.41 91 18.5 101 4.23 84 2.92 73 3.91 61 2.82 66 3.45 67 9.14 75 3.27 63
F-TV-L1 [15]77.9 5.44 100 12.5 82 5.69 104 5.46 98 15.0 99 4.03 93 7.48 88 16.3 87 3.42 76 5.08 94 23.3 100 2.81 88 3.42 73 4.34 70 3.03 76 4.05 67 15.1 81 3.18 53 2.43 46 3.92 62 1.87 30 3.90 73 9.35 79 2.61 50
Local-TV-L1 [65]78.5 5.33 98 12.6 83 5.19 102 6.90 104 15.7 102 6.22 102 10.0 104 18.2 97 8.89 97 5.81 100 24.7 105 3.70 99 3.05 44 4.00 46 2.39 39 4.05 67 14.6 77 3.09 50 1.95 17 3.11 4 2.15 44 5.85 100 10.8 95 7.34 107
SuperFlow [81]78.5 4.16 63 11.1 66 3.32 62 4.80 95 12.2 71 4.68 97 7.80 92 16.0 86 10.6 105 5.16 96 22.4 95 3.24 96 3.39 72 4.24 66 3.71 96 3.44 43 13.7 68 2.91 44 3.19 84 4.62 86 1.87 30 4.74 90 10.6 94 4.24 88
DF-Auto [115]78.8 5.04 96 13.7 92 3.30 58 6.51 101 14.1 93 6.09 101 8.14 97 16.5 89 10.2 103 5.06 93 21.3 88 3.10 95 3.74 90 4.91 105 3.25 83 2.67 11 11.4 41 2.14 7 3.36 86 5.23 106 1.45 7 4.45 83 9.18 76 4.28 89
TriFlow [95]79.1 4.73 94 12.4 78 3.49 78 4.03 81 12.5 76 3.70 89 8.18 99 17.2 92 10.4 104 3.50 59 15.4 42 2.32 79 3.43 75 4.21 60 3.42 87 3.90 61 12.3 52 3.76 73 7.86 127 5.72 113 16.2 127 2.80 57 5.89 51 2.50 45
CLG-TV [48]79.5 4.00 56 10.3 58 3.40 72 4.33 88 12.3 72 4.08 94 6.78 83 15.5 79 3.64 84 4.07 78 17.7 65 2.39 82 3.79 94 4.86 100 3.23 82 4.48 76 16.5 92 3.80 75 3.55 93 4.65 87 2.89 74 4.00 76 10.1 86 3.18 62
CBF [12]81.0 3.88 51 10.2 55 3.50 79 4.60 92 11.3 65 5.06 98 5.43 69 13.1 64 3.39 75 4.09 79 21.2 87 2.16 71 3.80 97 4.72 96 3.52 93 4.33 73 14.4 74 3.01 48 4.97 119 5.51 111 4.93 111 3.99 75 9.27 78 3.91 84
Bartels [41]82.2 4.43 75 11.1 66 4.17 94 2.83 37 8.84 29 2.56 60 4.54 56 12.5 60 2.80 63 4.87 89 22.1 93 3.05 93 3.58 85 4.35 74 4.15 105 5.55 93 17.5 96 5.78 101 3.74 100 5.02 96 5.98 116 5.21 96 11.9 101 5.20 96
Fusion [6]82.5 4.43 75 13.7 92 4.08 91 2.47 22 8.91 30 2.24 34 3.70 34 9.68 33 3.12 69 3.68 69 19.8 74 2.54 85 4.26 110 5.16 111 4.31 108 6.32 101 16.8 93 6.15 105 4.55 115 5.78 115 3.10 82 7.12 110 13.6 111 7.86 111
p-harmonic [29]83.2 4.64 90 13.0 86 4.43 98 3.41 74 11.9 68 2.93 79 7.60 89 18.1 96 3.96 88 4.65 87 21.0 85 2.97 91 3.46 77 4.33 69 3.34 85 4.75 83 17.5 96 4.60 92 3.05 78 4.17 71 2.15 44 5.09 95 10.9 96 3.77 80
CNN-flow-warp+ref [117]84.2 4.93 95 14.5 104 4.29 96 4.18 84 11.9 68 4.24 95 8.23 100 19.7 105 6.35 94 5.13 95 24.4 104 2.96 89 3.55 84 4.40 76 3.85 98 3.82 57 15.0 79 3.39 62 1.96 18 3.44 23 2.14 43 10.0 119 14.8 116 10.8 117
Dynamic MRF [7]84.8 4.58 85 12.4 78 4.14 92 3.25 64 13.9 90 2.27 38 6.02 76 16.8 90 2.36 41 4.39 83 22.6 96 2.51 84 3.61 87 4.55 85 3.46 89 6.81 106 22.2 116 6.78 110 2.41 44 3.48 29 3.69 100 9.26 117 17.8 120 10.2 114
SegOF [10]85.2 5.85 101 13.5 89 3.98 89 7.40 105 14.9 97 8.13 110 8.55 102 17.3 93 9.01 98 6.50 104 18.1 68 5.14 106 3.90 101 4.53 83 4.81 113 6.57 105 21.7 114 6.81 111 1.65 3 3.49 31 1.08 2 3.71 71 9.23 77 3.63 73
FlowNetS+ft+v [112]86.0 4.22 68 12.1 77 3.48 77 4.50 91 13.4 86 3.85 92 8.29 101 18.4 99 6.20 93 4.87 89 21.6 90 3.01 92 3.93 102 5.04 108 3.47 91 3.71 54 15.3 83 3.21 55 3.32 85 5.12 99 3.87 102 3.76 72 9.44 80 3.74 79
LDOF [28]86.8 4.60 86 13.0 86 3.77 86 4.67 93 15.5 101 3.67 88 5.63 72 14.0 69 4.21 89 5.80 99 27.1 114 3.43 97 3.52 82 4.50 82 3.46 89 4.84 86 17.8 98 4.04 79 2.46 50 4.14 70 3.25 90 4.85 93 12.0 102 3.78 81
Second-order prior [8]86.8 4.03 57 11.6 72 3.35 64 3.88 79 14.0 92 3.08 83 7.21 86 17.6 95 3.57 81 4.14 80 19.9 76 2.31 78 3.66 88 4.86 100 2.73 68 7.32 109 21.2 112 6.76 109 4.02 103 4.58 85 4.01 104 4.27 79 10.4 90 5.12 93
FlowNet2 [122]90.2 8.58 115 18.6 113 6.31 106 9.39 112 17.6 107 9.09 113 8.06 96 15.8 83 9.81 101 5.61 98 16.2 57 4.12 101 4.04 105 4.88 102 3.79 97 4.92 87 16.2 90 4.50 90 4.28 110 6.73 122 2.84 68 2.05 39 4.54 38 1.41 16
StereoFlow [44]90.4 17.1 129 28.1 129 17.9 128 18.7 126 29.7 127 16.5 121 20.1 126 30.9 126 17.5 123 21.2 127 38.3 128 17.9 125 4.60 114 5.05 109 5.52 115 2.38 4 11.5 45 1.77 2 1.25 1 2.92 2 0.71 1 4.49 85 10.3 89 4.23 87
Ad-TV-NDC [36]92.0 8.36 114 14.0 97 11.1 122 12.9 119 19.9 113 12.8 119 14.4 115 23.1 108 12.1 109 7.40 107 20.6 84 6.33 107 3.47 79 4.66 92 2.39 39 3.95 64 13.8 69 3.51 65 2.48 51 3.75 50 2.05 40 9.75 118 12.1 104 16.7 124
Learning Flow [11]95.0 4.23 69 11.7 73 3.41 74 4.16 83 15.3 100 3.42 86 6.78 83 16.9 91 3.83 86 6.41 103 25.3 108 4.25 102 4.66 116 6.01 123 4.00 101 6.33 103 20.7 111 5.30 95 3.09 81 4.84 90 2.91 77 7.08 109 15.0 117 5.27 98
Shiralkar [42]95.2 4.64 90 14.1 100 3.94 87 4.29 87 16.9 105 2.77 73 7.75 90 18.8 101 3.19 72 5.54 97 25.0 107 3.56 98 3.51 81 4.55 85 3.04 77 7.41 110 20.1 110 6.41 106 3.76 101 4.35 77 5.28 112 6.56 106 14.4 115 5.30 99
StereoOF-V1MT [119]95.8 4.71 93 14.1 100 3.95 88 5.10 97 20.3 115 2.78 74 7.98 95 20.7 106 2.57 53 4.48 85 21.1 86 2.79 87 4.20 109 5.29 113 4.10 103 6.85 108 22.3 117 6.42 107 2.45 49 4.17 71 3.15 86 10.5 120 18.4 123 10.5 115
IAOF2 [51]96.6 5.38 99 13.7 92 4.50 99 5.95 100 14.6 95 5.61 100 8.80 103 18.8 101 9.40 99 12.2 117 23.8 103 13.1 120 3.86 98 4.89 103 3.12 80 5.21 90 14.9 78 4.54 91 4.33 111 5.15 100 3.93 103 4.39 81 8.57 72 3.87 83
Modified CLG [34]98.0 7.17 109 17.1 112 6.47 108 6.85 103 14.9 97 7.48 106 14.0 111 24.8 112 15.7 119 8.35 110 27.3 115 6.36 108 3.96 103 4.99 107 4.08 102 4.54 77 19.3 106 4.15 83 2.33 41 3.86 59 2.40 51 6.00 101 13.8 113 5.40 100
2D-CLG [1]98.3 10.1 117 22.6 121 7.59 113 9.84 114 16.9 105 11.1 118 16.9 121 28.2 121 18.8 125 14.1 120 31.1 119 13.1 120 3.86 98 4.62 90 4.53 110 5.98 99 21.2 112 5.97 103 1.76 5 3.14 5 1.46 8 6.29 103 12.9 110 5.81 102
GraphCuts [14]99.0 6.25 102 14.3 103 5.53 103 8.60 108 20.1 114 6.61 104 7.91 94 15.4 78 10.9 106 4.88 91 19.0 70 3.05 93 3.78 92 4.71 94 3.94 99 8.74 118 16.4 91 5.39 97 4.04 104 4.87 92 4.85 110 6.35 104 12.2 105 6.05 103
Filter Flow [19]99.0 6.48 103 14.6 105 4.96 100 5.73 99 15.7 102 5.07 99 10.1 105 18.6 100 14.3 115 9.04 112 23.3 100 7.80 112 3.98 104 4.71 94 4.21 107 5.86 97 15.0 79 5.41 98 4.98 120 6.87 123 2.78 63 4.82 92 8.66 73 3.65 75
SPSA-learn [13]99.2 6.84 108 16.7 110 6.74 109 8.47 107 19.4 111 7.49 107 12.5 107 23.1 108 13.1 113 8.40 111 25.8 111 7.08 110 3.87 100 4.66 92 4.10 103 6.32 101 18.8 102 6.89 112 2.56 53 3.85 58 1.79 20 7.29 111 12.5 107 7.47 109
HBpMotionGpu [43]100.8 6.57 105 15.0 107 5.17 101 8.29 106 18.0 108 8.29 111 14.1 112 26.5 115 13.2 114 6.12 102 25.3 108 3.94 100 3.79 94 4.62 90 3.97 100 4.80 85 15.7 84 4.11 81 4.40 112 5.20 103 2.87 73 6.28 102 11.7 100 7.31 106
GroupFlow [9]101.3 8.00 111 18.6 113 8.09 115 11.1 117 23.7 120 10.3 116 12.6 108 25.6 113 12.8 111 5.84 101 20.3 82 4.39 103 4.69 117 5.81 119 3.67 94 9.29 119 22.4 118 10.1 121 2.11 30 3.99 66 2.29 48 5.75 98 10.0 84 7.39 108
IAOF [50]101.5 6.49 104 14.6 105 6.42 107 9.22 111 18.5 109 7.94 109 16.4 120 27.4 119 13.0 112 8.22 108 22.2 94 7.73 111 3.77 91 4.76 98 3.42 87 6.84 107 18.8 102 4.23 84 3.59 95 4.46 80 2.83 67 7.51 113 10.1 86 10.6 116
Black & Anandan [4]101.9 6.81 107 15.4 108 7.43 111 8.77 109 19.5 112 7.35 105 13.0 109 22.9 107 12.5 110 8.29 109 26.1 112 6.77 109 4.18 108 5.28 112 3.69 95 6.19 100 20.0 109 5.34 96 3.63 96 5.05 97 1.79 20 6.45 105 12.2 105 5.17 95
BlockOverlap [61]105.1 6.67 106 13.1 88 5.87 105 6.62 102 13.9 90 6.53 103 10.6 106 19.5 104 10.1 102 6.97 106 24.9 106 5.13 105 4.38 111 4.61 89 6.37 122 7.47 112 15.7 84 6.05 104 6.23 123 6.41 121 13.0 126 6.92 108 9.60 81 12.2 119
Nguyen [33]105.2 7.88 110 16.8 111 7.02 110 13.4 120 19.0 110 15.3 120 17.6 122 28.9 122 17.2 122 12.0 116 26.9 113 11.6 118 4.38 111 5.07 110 5.58 118 5.69 95 19.7 108 5.93 102 2.75 65 4.02 68 1.91 35 6.59 107 12.5 107 6.52 105
UnFlow [129]105.9 14.6 127 25.8 126 9.09 119 9.40 113 16.8 104 9.89 115 14.2 113 26.9 116 11.2 107 10.0 113 25.4 110 8.67 114 5.43 123 5.90 120 6.72 123 8.64 116 24.0 120 9.41 119 3.51 91 4.90 93 1.37 6 4.37 80 12.6 109 3.33 65
2bit-BM-tele [98]107.5 8.00 111 15.8 109 8.40 117 4.91 96 13.4 86 4.67 96 8.14 97 19.0 103 5.12 92 6.62 105 23.5 102 5.04 104 4.08 106 4.78 99 4.61 112 8.68 117 18.8 102 8.31 115 6.46 125 7.08 125 9.47 122 7.36 112 14.1 114 9.62 113
Horn & Schunck [3]110.9 8.01 113 19.9 115 8.38 116 9.13 110 23.2 119 7.71 108 14.2 113 25.9 114 14.6 117 12.4 118 30.6 117 11.3 117 4.64 115 5.64 115 4.60 111 8.21 115 24.4 121 8.45 116 4.01 102 5.41 107 1.95 37 9.16 116 17.5 118 8.86 112
SILK [79]112.2 9.34 116 20.4 116 10.5 121 10.4 115 21.9 116 10.3 116 16.0 119 27.5 120 14.5 116 10.3 114 29.0 116 8.54 113 4.81 118 5.65 116 5.56 117 9.41 120 25.4 123 8.74 117 2.79 68 3.68 48 4.62 108 10.9 121 17.8 120 12.3 120
Heeger++ [104]113.5 11.9 122 21.8 119 8.08 114 12.5 118 29.7 127 9.42 114 14.8 116 27.1 117 9.68 100 14.3 121 31.0 118 12.7 119 4.98 120 5.74 117 4.97 114 17.5 127 34.1 128 18.4 127 2.75 65 5.44 108 2.15 44 12.3 123 18.8 124 14.8 122
TI-DOFE [24]114.4 13.4 125 23.2 122 16.5 127 16.5 123 24.1 121 18.2 125 20.2 127 31.1 127 20.6 126 19.9 126 32.9 122 20.8 127 4.89 119 5.90 120 5.54 116 8.04 114 23.9 119 8.81 118 2.97 75 4.34 76 1.88 33 10.9 121 17.7 119 11.9 118
HCIC-L [99]118.1 15.7 128 22.0 120 10.1 120 31.5 129 26.6 125 41.0 129 14.8 116 23.1 108 16.8 121 18.4 125 34.4 124 18.2 126 5.94 124 6.35 124 6.35 121 10.6 123 19.2 105 11.4 123 18.7 129 17.8 129 19.2 128 4.93 94 8.34 70 5.16 94
SLK [47]118.2 11.6 120 26.0 127 14.6 126 15.3 122 25.0 123 17.5 123 17.8 124 30.1 125 18.1 124 25.4 129 33.6 123 28.0 129 5.25 121 5.90 120 7.03 124 10.3 122 27.4 125 10.6 122 2.89 72 4.47 81 2.94 79 14.9 126 20.7 126 18.8 125
FFV1MT [106]119.2 12.0 123 23.3 123 8.83 118 10.7 116 26.6 125 8.71 112 15.6 118 29.0 123 12.0 108 16.6 124 36.3 127 15.5 123 6.51 127 6.40 125 10.4 127 16.2 126 30.7 127 17.7 126 3.41 89 5.44 108 3.35 96 12.3 123 18.8 124 14.8 122
Adaptive flow [45]120.8 13.2 124 20.8 117 14.0 125 17.1 125 22.0 117 17.9 124 18.1 125 27.1 117 22.8 128 11.8 115 31.1 119 10.5 115 6.35 126 7.13 127 6.25 120 9.87 121 21.8 115 9.44 120 12.6 128 11.4 128 20.0 129 7.75 114 13.6 111 7.73 110
PGAM+LK [55]122.0 11.8 121 25.6 124 13.9 124 14.8 121 24.4 122 16.7 122 13.2 110 24.0 111 15.0 118 16.2 123 41.2 129 15.3 122 5.40 122 5.45 114 8.10 125 12.3 125 26.5 124 12.1 124 7.42 126 8.24 127 7.87 120 13.2 125 18.3 122 19.4 126
Periodicity [78]122.8 11.2 119 27.0 128 7.46 112 16.6 124 29.8 129 18.2 125 25.3 129 31.2 129 24.9 129 12.7 119 35.7 126 11.1 116 31.7 129 41.4 129 25.1 129 23.8 129 41.5 129 23.8 129 2.92 73 5.62 112 6.90 118 18.6 128 33.1 129 22.3 127
FOLKI [16]123.6 10.5 118 25.6 124 11.9 123 20.9 127 26.2 124 26.1 127 17.6 122 31.1 127 16.5 120 15.4 122 32.6 121 16.0 124 6.16 125 6.53 126 9.07 126 12.2 124 29.7 126 13.0 125 4.67 117 5.83 116 9.41 121 18.2 127 22.8 127 25.1 128
Pyramid LK [2]126.1 13.9 126 20.9 118 21.4 129 24.1 128 23.1 118 30.2 128 20.9 128 29.5 124 21.9 127 22.2 128 34.6 125 25.0 128 18.7 128 23.1 128 20.2 128 21.2 128 24.5 122 21.0 128 6.41 124 7.02 124 10.8 124 25.6 129 31.5 128 34.5 129
AdaConv-v1 [126]130.0 39.2 130 39.9 130 41.8 130 73.0 130 74.5 130 71.1 130 70.1 130 67.3 130 71.8 130 64.4 130 66.2 130 65.9 130 76.5 130 78.1 130 72.0 130 68.2 130 64.9 130 66.5 130 52.3 130 45.1 130 70.9 130 81.8 130 81.6 130 82.3 130
SepConv-v1 [127]130.0 39.2 130 39.9 130 41.8 130 73.0 130 74.5 130 71.1 130 70.1 130 67.3 130 71.8 130 64.4 130 66.2 130 65.9 130 76.5 130 78.1 130 72.0 130 68.2 130 64.9 130 66.5 130 52.3 130 45.1 130 70.9 130 81.8 130 81.6 130 82.3 130
Move the mouse over the numbers in the table to see the corresponding images. Click to compare with the ground truth.

References

Methodtime*framescolor Reference and notes
[1] 2D-CLG 844 2 gray The 2D-CLG method by Bruhn et al. as implemented by Stefan Roth. [A. Bruhn, J. Weickert, and C. Schnörr. Lucas/Kanade meets Horn/Schunck: combining local and global optic flow methods. IJCV 63(3), 2005.] Parameters were set to match the published performance on Yosemite sequence, which may not be optimal for other sequences.
[2] Pyramid LK 12 2 color A modification of Bouguet's pyramidal implementation of Lucas-Kanade.
[3] Horn & Schunck 49 2 gray A modern Matlab implementation of the Horn & Schunck method by Deqing Sun. Parameters set to optimize AAE on all training data.
[4] Black & Anandan 328 2 gray A modern Matlab implementation of the Black & Anandan method by Deqing Sun.
[5] Brox et al. 18 2 color T. Brox, A. Bruhn, N. Papenberg, and J. Weickert. High accuracy optical flow estimation based on a theory for warping. ECCV 2004. (Improved using separate robust functions as proposed in A. Bruhn and J. Weickert, Towards ultimate motion estimation, ICCV 2005; improved by training on the training set.)
[6] Fusion 2,666 2 color V. Lempitsky, S. Roth, and C. Rother. Discrete-continuous optimization for optical flow estimation. CVPR 2008.
[7] Dynamic MRF 366 2 gray B. Glocker, N. Paragios, N. Komodakis, G. Tziritas, and N. Navab. Optical flow estimation with uncertainties through dynamic MRFs. CVPR 2008. (Method improved since publication.)
[8] Second-order prior 14 2 gray W. Trobin, T. Pock, D. Cremers, and H. Bischof. An unbiased second-order prior for high-accuracy motion estimation. DAGM 2008. (Method improved since publication; for details see W. Trobin, Ph.D. thesis, 2009.)
[9] GroupFlow 600 2 gray X. Ren. Local Grouping for Optical Flow. CVPR 2008.
[10] SegOF 60 2 color L. Xu, J. Chen, and J. Jia. Segmentation based variational model for accurate optical flow estimation. ECCV 2008. Code available.
[11] Learning Flow 825 2 gray D. Sun, S. Roth, J.P. Lewis, and M. Black. Learning optical flow (SRF-LFC). ECCV 2008.
[12] CBF 69 2 color W. Trobin, T. Pock, D. Cremers, and H. Bischof. Continuous energy minimization via repeated binary fusion. ECCV 2008. (Method improved since publication; for details see W. Trobin, Ph.D. thesis, 2009.)
[13] SPSA-learn 200 2 color Y. Li and D. Huttenlocher. Learning for optical flow using stochastic optimization. ECCV 2008.
[14] GraphCuts 1,200 2 color T. Cooke. Two applications of graph-cuts to image processing. DICTA 2008.
[15] F-TV-L1 8 2 gray A. Wedel, T. Pock, J. Braun, U. Franke, and D. Cremers. Duality TV-L1 flow with fundamental matrix prior. IVCNZ 2008.
[16] FOLKI 1.4 2 gray G. Le Besnerais and F. Champagnat. Dense optical flow by iterative local window registration. ICIP 2005.
[17] TV-L1-improved 2.9 2 gray A. Wedel, T. Pock, C. Zach, H. Bischof, and D. Cremers. An improved algorithm for TV-L1 optical flow computation. Proceedings of the Dagstuhl Visual Motion Analysis Workshop 2008. Code at GPU4Vision.
[18] DPOF 287 2 color C. Lei and Y.-H. Yang. Optical flow estimation on coarse-to-fine region-trees using discrete optimization. ICCV 2009. (Method improved since publication.)
[19] Filter Flow 34,000 2 color S. Seitz and S. Baker. Filter flow. ICCV 2009.
[20] Adaptive 9.2 2 gray A. Wedel, D. Cremers, T. Pock, and H. Bischof. Structure- and motion-adaptive regularization for high accuracy optic flow. ICCV 2009.
[21] Complementary OF 44 2 color H. Zimmer, A. Bruhn, J. Weickert, L. Valgaerts, A. Salgado, B. Rosenhahn, and H.-P. Seidel. Complementary optic flow. EMMCVPR 2009.
[22] Aniso. Huber-L1 2 2 gray M. Werlberger, W. Trobin, T. Pock, A. Wedel, D. Cremers, and H. Bischof. Anisotropic Huber-L1 optical flow. BMVC 2009. Code at GPU4Vision.
[23] Rannacher 0.12 2 gray J. Rannacher. Realtime 3D motion estimation on graphics hardware. Bachelor thesis, Heidelberg University, 2009.
[24] TI-DOFE 260 2 gray C. Cassisa, S. Simoens, and V. Prinet. Two-frame optical flow formulation in an unwarped multiresolution scheme. CIARP 2009.
[25] NL-TV-NCC 20 2 color M. Werlberger, T. Pock, and H. Bischof. Motion estimation with non-local total variation regularization. CVPR 2010.
[26] MDP-Flow 188 2 color L. Xu, J. Jia, and Y. Matsushita. Motion detail preserving optical flow estimation. CVPR 2010.
[27] ACK-Prior 5872 2 color K. Lee, D. Kwon, I. Yun, and S. Lee. Optical flow estimation with adaptive convolution kernel prior on discrete framework. CVPR 2010.
[28] LDOF 122 2 color T. Brox and J. Malik. Large displacement optical flow: descriptor matching in variational motion estimation. PAMI 33(3):500-513, 2011.
[29] p-harmonic 565 2 gray J. Gai and R. Stevenson. Optical flow estimation with p-harmonic regularization. ICIP 2010.
[30] TriangleFlow 4200 2 gray B. Glocker, H. Heibel, N. Navab, P. Kohli, and C. Rother. TriangleFlow: Optical flow with triangulation-based higher-order likelihoods. ECCV 2010.
[31] Classic+NL 972 2 color D. Sun, S. Roth, and M. Black. Secrets of optical flow estimation and their principles. CVPR 2010. Matlab code.
[32] Classic++ 486 2 gray A modern implementation of the classical formulation descended from Horn & Schunck and Black & Anandan; see D. Sun, S. Roth, and M. Black, Secrets of optical flow estimation and their principles, CVPR 2010.
[33] Nguyen 33 2 gray D. Nguyen. Tuning optical flow estimation with image-driven functions. ICRA 2011.
[34] Modified CLG 133 2 gray R. Fezzani, F. Champagnat, and G. Le Besnerais. Combined local global method for optic flow computation. EUSIPCO 2010.
[35] ComplOF-FED-GPU 0.97 2 color P. Gwosdek, H. Zimmer, S. Grewenig, A. Bruhn, and J. Weickert. A highly efficient GPU implementation for variational optic flow based on the Euler-Lagrange framework. CVGPU Workshop 2010.
[36] Ad-TV-NDC 35 2 gray M. Nawaz. Motion estimation with adaptive regularization and neighborhood dependent constraint. DICTA 2010.
[37] Layers++ 18206 2 color D. Sun, E. Sudderth, and M. Black. Layered image motion with explicit occlusions, temporal consistency, and depth ordering. NIPS 2010.
[38] OFH 620 3 color H. Zimmer, A. Bruhn, J. Weickert. Optic flow in harmony. IJCV 93(3) 2011.
[39] LSM 1615 2 color K. Jia, X. Wang, and X. Tang. Optical flow estimation using learned sparse model. ICCV 2011.
[40] CostFilter 55 2 color C. Rhemann, A. Hosni, M. Bleyer, C. Rother, and M. Gelautz. Fast cost-volume filtering for visual correspondence and beyond. CVPR 2011.
[41] Bartels 0.15 2 gray C. Bartels and G. de Haan. Smoothness constraints in recursive search motion estimation for picture rate conversion. IEEE TCSVT 2010. Version improved since publication: mapped on GPU.
[42] Shiralkar 600 2 gray M. Shiralkar and R. Schalkoff. A self organization-based optical flow estimator with GPU implementation. MVA 23(6):1229-1242.
[43] HBpMotionGpu 1000 5 gray S. Grauer-Gray and C. Kambhamettu. Hierarchical belief propagation to reduce search space using CUDA for stereo and motion estimation. WACV 2009. (Method improved since publication.)
[44] StereoFlow 7200 2 color G. Rosman, S. Shem-Tov, D. Bitton, T. Nir, G. Adiv, R. Kimmel, A. Feuer, and A. Bruckstein. Over-parameterized optical flow using a stereoscopic constraint. SSVM 2011:761-772.
[45] Adaptive flow 121 2 gray T. Arici. Energy minimization based motion estimation using adaptive smoothness priors. Submitted to IEEE TIP 2011.
[46] TC-Flow 2500 5 color S. Volz, A. Bruhn, L. Valgaerts, and H. Zimmer. Modeling temporal coherence for optical flow. ICCV 2011.
[47] SLK 300 2 gray T. Corpetti and E. Mémin. Stochastic uncertainty models for the luminance consistency assumption. IEEE TIP 2011.
[48] CLG-TV 29 2 gray M. Drulea. Total variation regularization of local-global optical flow. ITSC 2011. Matlab code.
[49] SimpleFlow 1.7 2 color M. Tao, J. Bai, P. Kohli, S. Paris. SimpleFlow: a non-iterative, sublinear optical flow algorithm. EUROGRAPHICS 2012.
[50] IAOF 57 2 gray D. Nguyen. Improving motion estimation using image-driven functions and hybrid scheme. PSIVT 2011.
[51] IAOF2 56 2 gray D. Nguyen. Enhancing the sharpness of flow field using image-driven functions with occlusion-aware filter. Submitted to TIP 2011.
[52] LocallyOriented 9541 2 gray Y.Niu, A. Dick, and M. Brooks. Locally oriented optical flow computation. To appear in TIP 2012.
[53] IROF-TV 261 2 color H. Rashwan, D. Puig, and M. Garcia. On improving the robustness of differential optical flow. ICCV 2011 Artemis workshop.
[54] Sparse Occlusion 2312 2 color A. Ayvaci, M. Raptis, and S. Soatto. Sparse occlusion detection with optical flow. Submitted to IJCV 2011.
[55] PGAM+LK 0.37 2 gray A. Alba, E. Arce-Santana, and M. Rivera. Optical flow estimation with prior models obtained from phase correlation. ISVC 2010.
[56] Sparse-NonSparse 713 2 color L. Chen, J. Wang, and Y. Wu. Decomposing and regularizing sparse/non-sparse components for motion field estimation. Submitted to PAMI 2013.
[57] nLayers 36150 4 color D. Sun, E. Sudderth, and M. Black. Layered segmentation and optical flow estimation over time. CVPR 2012.
[58] IROF++ 187 2 color H. Rashwan, D. Puig, and M. Garcia. Variational optical flow estimation based on stick tensor voting. IEEE TIP 2013.
[59] COFM 600 3 color M. Mozerov. Constrained optical flow estimation as a matching problem. IEEE TIP 2013.
[60] Efficient-NL 400 2 color P. Krähenbühl and V. Koltun. Efficient nonlocal regularization for optical flow. ECCV 2012.
[61] BlockOverlap 2 2 gray M. Santoro, G. AlRegib, and Y. Altunbasak. Motion estimation using block overlap minimization. Submitted to MMSP 2012.
[62] Ramp 1200 2 color A. Singh and N. Ahuja. Exploiting ramp structures for improving optical flow estimation. ICPR 2012.
[63] Occlusion-TV-L1 538 3 gray C. Ballester, L. Garrido, V. Lazcano, and V. Caselles. A TV-L1 optical flow method with occlusion detection. DAGM-OAGM 2012.
[64] TV-L1-MCT 90 2 color M. Mohamed and B. Mertsching. TV-L1 optical flow estimation with image details recovering based on modified census transform. ISVC 2012.
[65] Local-TV-L1 500 2 gray L. Raket. Local smoothness for global optical flow. ICIP 2012.
[66] ALD-Flow 61 2 color M. Stoll, A. Bruhn, and S. Volz. Adaptive integration of feature matches into variational optic flow methods. ACCV 2012.
[67] SIOF 234 2 color L. Xu, Z. Dai, and J. Jia. Scale invariant optical flow. ECCV 2012.
[68] MDP-Flow2 342 2 color L. Xu, J. Jia, and Y. Matsushita. Motion detail preserving optical flow estimation. PAMI 34(9):1744-1757, 2012. Code available.
[69] TCOF 1421 all gray J. Sanchez, A. Salgado, and N. Monzon. Optical flow estimation with consistent spatio-temporal coherence models. VISAPP 2013.
[70] LME 476 2 color W. Li, D. Cosker, M. Brown, and R. Tang. Optical flow estimation using Laplacian mesh energy. CVPR 2013.
[71] NN-field 362 2 color L. Chen, H. Jin, Z. Lin, S. Cohen, and Y. Wu. Large displacement optical flow from nearest neighbor fields. CVPR 2013.
[72] FESL 3310 2 color W. Dong, G. Shi, X. Hu, and Y. Ma. Nonlocal sparse and low-rank regularization for optical flow estimation. Submitted to IEEE TIP 2013.
[73] PMF 35 2 color J. Lu, H. Yang, D. Min, and M. Do. PatchMatch filter: efficient edge-aware filtering meets randomized search for fast correspondence field estimation. CVPR 2013.
[74] FC-2Layers-FF 2662 4 color D. Sun, J. Wulff, E. Sudderth, H. Pfister, and M. Black. A fully-connected layered model of foreground and background flow. CVPR 2013.
[75] Correlation Flow 290 2 color M. Drulea and S. Nedevschi. Motion estimation using the correlation transform. TIP 2013. Matlab code.
[76] TC/T-Flow 341 5 color M. Stoll, S. Volz, and A. Bruhn. Joint trilateral filtering for multiframe optical flow. ICIP 2013.
[77] OFLAF 1530 2 color T. Kim, H. Lee, and K. Lee. Optical flow via locally adaptive fusion of complementary data costs. ICCV 2013.
[78] Periodicity 8000 4 color G. Khachaturov, S. Gonzalez-Brambila, and J. Gonzalez-Trejo. Periodicity-based computation of optical flow. Submitted to Computacion y Sistemas (CyS) 2013.
[79] SILK 572 2 gray P. Zille, C. Xu, T. Corpetti, L. Shao. Observation models based on scale interactions for optical flow estimation. Submitted to IEEE TIP.
[80] CRTflow 13 3 color O. Demetz, D. Hafner, and J. Weickert. The complete rank transform: a tool for accurate and morphologically invariant matching of structures. BMVC 2013.
[81] SuperFlow 178 2 color Anonymous. Superpixel based optical flow estimation. ICCV 2013 submission 507.
[82] Aniso-Texture 300 2 color Anonymous. Texture information-based optical flow estimation using an incremental multi-resolution approach. ITC-CSCC 2013 submission 267.
[83] Classic+CPF 640 2 gray Z. Tu, R. Veltkamp, and N. van der Aa. A combined post-filtering method to improve accuracy of variational optical flow estimation. Submitted to Pattern Recognition 2013.
[84] S2D-Matching 1200 2 color Anonymous. Locally affine sparse-to-dense matching for motion and occlusion estimation. ICCV 2013 submission 1479.
[85] AGIF+OF 438 2 gray Z. Tu, R. Poppe, and R. Veltkamp. Adaptive guided image filter to warped interpolation image for variational optical flow computation. Submitted to Signal Processing 2015.
[86] DeepFlow 13 2 color P. Weinzaepfel, J. Revaud, Z. Harchaoui, and C. Schmid. DeepFlow: large displacement optical flow with deep matching. ICCV 2013.
[87] NNF-Local 673 2 color Z. Chen, H. Jin, Z. Lin, S. Cohen, and Y. Wu. Large displacement optical flow with nearest neighbor field. Submitted to PAMI 2014.
[88] EPPM w/o HM 2.5 2 color L. Bao, Q. Yang, and H. Jin. Fast edge-preserving PatchMatch for large displacement optical flow. CVPR 2014.
[89] MLDP_OF 165 2 gray M. Mohamed, H. Rashwan, B. Mertsching, M. Garcia, and D. Puig. Illumination-robust optical flow approach using local directional pattern. IEEE TCSVT 24(9):1499-1508, 2014.
[90] RFlow 20 2 gray S. Ali, C. Daul, and W. Blondel. Robust and accurate optical flow estimation for weak texture and varying illumination condition: Application to cystoscopy. IPTA 2014.
[91] SRR-TVOF-NL 32 all color P. Pohl, M. Sirotenko, E. Tolstaya, and V. Bucha. Edge preserving motion estimation with occlusions correction for assisted 2D to 3D conversion. IS&T/SPIE Electronic Imaging 2014.
[92] 2DHMM-SAS 157 2 color M.-C. Shih, R. Shenoy, and K. Rose. A two-dimensional hidden Markov model with spatially-adaptive states with application of optical flow. ICIP 2014 submission.
[93] WLIF-Flow 700 2 color Z. Tu, R. Veltkamp, N. van der Aa, and C. Van Gemeren. Weighted local intensity fusion method for variational optical flow estimation. Submitted to TIP 2014.
[94] FMOF 215 2 color N. Jith, A. Ramakanth, and V. Babu. Optical flow estimation using approximate nearest neighbor field fusion. ICASSP 2014.
[95] TriFlow 150 2 color TriFlow. Optical flow with geometric occlusion estimation and fusion of multiple frames. ECCV 2014 submission 914.
[96] ComponentFusion 6.5 2 color Anonymous. Fast optical flow by component fusion. ECCV 2014 submission 941.
[97] AggregFlow 1642 2 color D. Fortun, P. Bouthemy, and C. Kervrann. Aggregation of local parametric candidates and exemplar-based occlusion handling for optical flow. Preprint arXiv:1407.5759.
[98] 2bit-BM-tele 124 2 gray R. Xu and D. Taubman. Robust dense block-based motion estimation using a two-bit transform on a Laplacian pyramid. ICIP 2013.
[99] HCIC-L 330 2 color Anonymous. Globally-optimal image correspondence using a hierarchical graphical model. NIPS 2014 submission 114.
[100] TF+OM 600 2 color R. Kennedy and C. Taylor. Optical flow with geometric occlusion estimation and fusion of multiple frames. EMMCVPR 2015.
[101] PH-Flow 800 2 color J. Yang and H. Li. Dense, accurate optical flow estimation with piecewise parametric model. CVPR 2015.
[102] EpicFlow 16 2 color J. Revaud, P. Weinzaepfel, Z. Harchaoui, and C. Schmid. EpicFlow: edge-preserving interpolation of correspondences for optical flow. CVPR 2015.
[103] NNF-EAC 380 2 color Anonymous. Variational method for joint optical flow estimation and edge-aware image restoration. CVPR 2015 submission 2336.
[104] Heeger++ 6600 5 gray Anonymous. A context aware biologically inspired algorithm for optical flow (updated results). CVPR 2015 submission 2238.
[105] HBM-GC 330 2 color A. Zheng and Y. Yuan. Motion estimation via hierarchical block matching and graph cut. Submitted to ICIP 2015.
[106] FFV1MT 358 5 gray F. Solari, M. Chessa, N. Medathati, and P. Kornprobst. What can we expect from a V1-MT feedforward architecture for optical flow estimation? Submitted to Signal Processing: Image Communication 2015.
[107] ROF-ND 4 2 color S. Ali, C. Daul, E. Galbrun, and W. Blondel. Illumination invariant large displacement optical flow using robust neighbourhood descriptors. Submitted to CVIU 2015.
[108] DeepFlow2 16 2 color J. Revaud, P. Weinzaepfel, Z. Harchaoui, and C. Schmid. Deep convolutional matching. Submitted to IJCV, 2015.
[109] HAST 2667 2 color Anonymous. Highly accurate optical flow estimation on superpixel tree. ICCV 2015 submission 2221.
[110] FlowFields 15 2 color 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.
* The "time" column lists the reported runtime in seconds on the "Urban" sequence. Note that these runtimes are not normalized by processor speed or type.