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        
R10.0
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]8.5 3.83 4 16.9 9 1.87 7 2.64 7 16.1 11 1.33 7 3.02 3 10.7 3 1.33 13 2.79 15 18.9 21 1.11 16 4.82 1 6.85 1 1.73 2 4.13 4 14.6 3 2.27 2 0.52 19 5.08 47 0.00 1 0.28 3 1.04 3 0.04 2
OFLAF [77]9.0 3.84 5 17.2 11 1.99 12 2.48 5 14.1 5 1.41 8 2.96 2 10.1 2 1.17 9 2.36 8 14.7 7 0.82 9 5.15 3 7.90 7 2.15 4 6.43 31 18.2 12 4.98 18 0.27 2 2.42 4 0.07 15 0.79 9 1.88 7 1.81 21
MDP-Flow2 [68]10.7 3.86 7 17.2 11 1.90 9 2.06 1 12.6 1 1.04 2 3.22 5 11.0 4 1.16 8 3.27 26 21.7 31 1.19 20 6.35 16 8.86 13 3.12 10 5.40 13 15.7 4 5.11 20 0.38 11 3.75 24 0.02 3 0.49 5 1.80 6 0.13 6
NN-field [71]12.2 4.31 17 18.6 22 2.22 21 3.13 15 18.3 25 1.79 18 3.16 4 11.1 5 1.40 14 2.08 5 16.7 10 0.78 7 5.28 4 7.44 3 2.25 5 2.53 1 8.92 1 0.92 1 0.89 37 6.39 70 0.02 3 0.27 2 0.98 2 0.04 2
PMMST [114]13.5 4.02 10 16.5 7 1.46 1 3.86 31 16.9 17 3.33 45 3.91 10 12.6 9 2.82 38 2.18 7 9.47 1 1.30 26 5.68 9 7.88 6 2.78 7 5.25 11 13.7 2 4.29 10 0.53 21 5.11 48 0.02 3 0.26 1 0.94 1 0.04 2
nLayers [57]22.0 4.08 13 16.2 5 2.80 37 4.71 54 19.3 32 3.82 71 4.64 18 15.2 18 3.96 58 1.99 3 13.2 2 0.80 8 5.34 6 7.57 5 3.22 12 5.85 25 16.8 7 4.64 13 0.87 35 3.68 22 0.96 48 0.84 10 2.94 17 0.75 8
ComponentFusion [96]22.8 4.03 11 17.7 15 2.19 20 2.17 4 13.1 3 1.26 5 3.86 9 13.2 11 1.58 16 2.85 18 18.0 14 1.03 14 6.68 23 9.59 22 4.14 32 8.35 67 27.3 67 8.34 79 0.60 23 4.03 29 0.52 37 0.76 8 2.22 11 1.09 10
LME [70]23.2 3.70 3 16.1 4 1.69 2 2.13 2 13.0 2 1.19 4 5.91 44 15.4 19 7.43 83 3.23 22 22.4 35 1.19 20 6.60 22 9.12 16 4.39 41 6.11 27 20.8 21 6.60 49 0.52 19 4.96 45 0.07 15 1.09 14 3.28 24 1.86 23
SVFilterOh [111]23.8 4.36 19 15.9 3 2.01 15 3.04 13 16.3 13 1.78 17 3.31 6 11.3 6 1.20 10 2.02 4 13.6 4 0.57 2 5.94 12 8.69 12 2.12 3 6.61 33 19.3 14 6.32 45 5.10 95 12.9 108 10.0 107 0.75 7 2.20 10 1.32 12
FC-2Layers-FF [74]24.3 4.03 11 16.3 6 2.39 28 4.23 40 20.9 41 3.21 42 3.40 8 11.4 7 2.64 32 2.74 13 17.1 11 1.03 14 5.73 10 8.29 10 3.31 14 7.49 45 20.5 20 6.66 53 1.30 52 6.84 74 0.34 31 0.64 6 1.78 5 1.20 11
NNF-EAC [103]25.1 4.32 18 18.6 22 2.18 19 2.69 8 15.1 7 1.64 13 3.93 11 13.1 10 1.27 11 4.17 58 23.0 44 1.95 55 7.09 31 9.97 29 3.89 28 6.33 29 17.4 9 5.53 22 0.55 22 5.18 49 0.02 3 1.60 39 4.32 40 1.93 26
HAST [109]26.4 2.98 1 12.9 1 1.71 3 3.63 28 15.0 6 2.78 34 2.46 1 8.38 1 0.25 1 2.84 17 18.0 14 0.67 4 5.02 2 7.35 2 1.66 1 8.83 75 22.4 35 8.37 80 6.13 101 12.0 103 18.0 114 0.31 4 1.13 4 0.03 1
WLIF-Flow [93]27.7 3.97 9 17.0 10 2.12 17 3.53 22 18.5 27 2.37 28 4.60 17 15.1 16 2.34 28 3.40 29 20.3 27 1.50 31 7.69 54 11.4 62 4.79 54 6.67 34 17.8 11 5.53 22 0.40 14 3.68 22 0.07 15 1.59 38 3.82 34 2.63 43
RNLOD-Flow [121]28.5 3.58 2 15.7 2 1.83 6 3.60 26 19.9 35 2.06 22 5.53 38 18.1 40 2.42 30 2.75 14 17.7 12 1.01 13 6.01 13 9.09 15 3.98 29 7.21 39 20.0 16 6.87 62 2.59 73 9.67 86 1.95 65 1.06 13 2.66 14 1.79 20
FESL [72]28.8 3.91 8 16.6 8 2.13 18 5.68 75 23.5 58 4.23 76 5.17 29 16.8 25 2.99 40 2.41 10 15.8 8 0.89 12 5.76 11 8.62 11 4.05 30 5.81 22 17.6 10 5.32 21 1.09 45 5.68 62 1.21 55 1.35 23 2.89 16 1.72 18
Layers++ [37]29.0 4.39 20 17.8 16 3.14 45 3.70 29 18.0 22 2.84 35 3.37 7 11.5 8 2.65 33 2.38 9 14.1 6 0.82 9 5.33 5 7.52 4 3.78 26 7.58 48 22.0 31 6.13 39 1.81 68 7.08 76 0.54 38 1.45 30 2.46 13 4.56 79
ALD-Flow [66]30.1 4.22 14 18.2 19 1.93 11 3.20 18 16.8 16 1.59 12 5.21 30 17.4 32 1.13 7 3.70 37 22.9 42 1.26 22 6.54 21 9.31 18 3.14 11 5.25 11 21.5 26 4.98 18 0.88 36 4.67 37 4.48 90 2.69 58 6.66 57 4.79 80
TC/T-Flow [76]31.0 4.57 24 20.6 30 2.00 13 3.45 21 18.7 28 1.52 10 4.30 13 14.3 12 0.67 3 3.97 53 23.1 45 1.80 46 6.35 16 9.50 21 3.36 17 4.48 6 15.7 4 4.78 15 1.30 52 6.94 75 5.07 93 2.08 50 5.10 48 2.83 48
PMF [73]31.3 4.65 26 18.3 21 2.34 24 3.37 19 18.2 24 1.92 20 4.20 12 14.6 14 1.01 6 3.24 24 18.6 18 1.11 16 5.50 7 8.09 8 2.43 6 6.99 37 24.8 50 6.27 43 6.87 104 17.3 119 8.77 106 0.91 11 2.06 8 2.06 29
Correlation Flow [75]31.6 4.57 24 20.5 29 1.87 7 2.71 9 16.2 12 1.16 3 5.74 42 17.9 38 0.66 2 1.91 1 13.5 3 0.85 11 8.00 62 12.0 71 4.57 44 8.69 70 23.8 44 8.93 85 0.84 32 4.74 38 0.96 48 1.38 26 3.81 33 1.91 25
Efficient-NL [60]32.7 4.24 15 17.4 13 2.24 22 4.30 41 21.8 44 2.75 32 5.26 33 16.9 27 2.67 34 3.43 31 21.0 29 1.74 41 6.02 14 9.16 17 3.31 14 7.84 57 21.7 28 6.26 42 1.36 55 6.79 73 1.03 51 1.48 33 3.08 20 1.77 19
AGIF+OF [85]33.1 4.39 20 18.0 18 2.80 37 5.10 65 23.7 61 3.70 66 5.06 25 16.8 25 3.11 41 3.29 27 20.1 25 1.45 30 6.45 19 9.31 18 4.62 46 6.77 36 19.7 15 5.86 31 0.37 10 3.60 20 0.25 28 1.79 45 3.68 30 3.12 57
TC-Flow [46]34.3 4.27 16 19.0 25 1.91 10 2.85 10 16.6 15 1.45 9 5.05 24 16.9 27 0.80 5 4.05 54 23.9 53 1.74 41 6.73 24 9.68 24 2.93 9 5.83 23 22.9 39 5.68 28 1.39 57 4.87 43 7.32 102 2.46 55 6.13 54 4.49 76
Classic+CPF [83]36.0 4.77 30 19.7 27 2.99 40 4.59 46 23.3 56 3.08 40 5.24 31 17.2 30 2.81 37 3.32 28 21.3 30 1.57 35 6.51 20 9.49 20 4.24 37 7.39 42 21.5 26 6.27 43 1.02 40 5.33 53 1.40 57 1.47 32 3.19 22 2.47 41
OAR-Flow [125]36.1 5.41 45 21.6 41 2.61 31 4.96 61 22.3 47 2.88 36 7.90 59 23.8 58 4.19 62 4.45 65 22.7 41 1.90 50 7.03 29 10.1 32 3.39 18 5.10 10 22.3 33 4.56 12 0.29 4 2.64 8 0.17 24 1.58 37 4.89 47 1.68 16
COFM [59]38.6 4.75 29 20.2 28 2.63 33 3.40 20 18.3 25 2.14 24 6.19 48 19.3 46 4.00 59 3.04 20 18.8 19 1.11 16 7.45 44 10.1 32 7.01 83 8.80 73 20.9 22 6.68 54 1.41 58 3.66 21 2.76 77 1.22 17 2.28 12 3.72 67
PH-Flow [101]38.8 5.11 37 21.0 32 3.62 59 4.59 46 22.4 48 3.37 49 4.37 14 14.5 13 3.45 46 3.93 50 22.5 37 2.07 62 6.34 15 9.00 14 3.74 24 7.28 41 21.7 28 6.39 47 1.61 63 5.58 60 1.58 60 1.15 16 2.12 9 3.39 62
HBM-GC [105]38.9 5.82 52 18.2 19 2.00 13 4.47 45 18.7 28 3.80 70 4.39 15 15.1 16 1.73 20 2.42 11 13.8 5 0.72 6 6.77 25 9.60 23 4.08 31 7.61 50 19.0 13 5.76 29 4.61 91 12.6 105 2.83 79 2.39 53 5.72 51 5.01 84
IROF++ [58]39.1 4.68 27 19.4 26 2.70 34 4.66 50 23.1 52 3.42 54 5.25 32 17.2 30 3.79 54 3.95 51 23.2 46 2.05 59 6.97 27 9.84 27 4.64 47 7.99 61 24.6 48 7.05 66 0.44 16 4.30 34 0.00 1 1.37 25 3.26 23 2.83 48
ProbFlowFields [128]39.5 8.29 81 31.1 80 5.73 93 3.54 23 18.0 22 2.75 32 6.07 47 18.9 45 5.22 65 3.54 34 17.7 12 1.91 51 7.66 53 10.8 49 4.59 45 5.06 9 20.0 16 5.58 26 0.38 11 3.08 14 0.07 15 1.70 42 4.56 43 2.41 40
CostFilter [40]40.6 5.29 43 22.0 43 2.85 39 3.54 23 17.7 20 2.16 25 4.64 18 16.0 20 1.75 21 3.68 36 22.5 37 1.27 24 5.67 8 8.14 9 2.85 8 7.76 54 25.9 57 6.80 58 6.98 107 24.2 124 12.9 109 1.43 27 4.11 37 2.02 28
Sparse-NonSparse [56]40.7 4.98 33 20.8 31 4.09 68 4.63 49 22.9 49 3.41 53 5.02 23 16.7 24 3.47 49 3.89 46 22.6 40 1.91 51 7.17 34 10.2 35 4.30 39 7.66 52 22.3 33 6.80 58 0.69 28 3.53 19 0.89 45 1.52 35 3.56 29 2.97 53
MLDP_OF [89]41.1 6.35 59 26.0 59 3.41 50 2.97 12 16.4 14 1.76 16 5.47 36 17.8 36 1.30 12 2.79 15 19.7 23 1.12 19 7.13 33 10.0 30 3.75 25 7.49 45 21.4 25 9.75 90 5.04 93 6.22 67 17.0 113 1.79 45 4.28 39 2.14 31
FMOF [94]41.8 4.42 22 17.8 16 3.06 43 5.03 62 23.1 52 3.63 62 4.45 16 14.8 15 2.80 36 2.94 19 18.8 19 1.26 22 7.00 28 10.2 35 4.71 50 8.92 76 20.9 22 7.13 68 1.06 43 6.34 69 1.85 64 2.58 57 5.80 52 3.06 55
Classic+NL [31]42.0 5.07 36 21.0 32 4.22 72 4.70 53 23.4 57 3.27 43 4.98 22 16.5 22 3.48 50 3.75 40 22.5 37 1.68 38 7.21 37 10.2 35 4.32 40 7.82 55 22.4 35 6.71 55 1.47 59 6.39 70 1.18 54 1.12 15 2.87 15 2.27 35
LSM [39]42.2 5.00 35 21.2 38 3.93 65 4.62 48 22.9 49 3.37 49 5.13 27 17.1 29 3.26 44 3.80 41 22.9 42 1.87 48 6.92 26 9.78 25 4.41 43 7.71 53 22.4 35 6.74 56 1.00 39 4.76 40 1.16 53 1.68 41 3.94 35 2.90 52
Ramp [62]42.6 5.12 38 21.1 36 3.82 64 4.68 52 23.2 54 3.47 57 4.89 21 16.3 21 3.46 47 3.83 43 22.3 34 1.93 54 7.23 38 10.2 35 4.80 55 7.61 50 22.1 32 6.80 58 1.20 48 5.04 46 1.43 58 1.36 24 2.98 18 2.31 39
Aniso-Texture [82]43.7 3.84 5 17.5 14 1.76 4 2.88 11 15.9 9 2.11 23 7.10 56 20.9 51 2.30 24 1.97 2 16.5 9 0.57 2 8.24 67 11.9 68 5.22 67 8.82 74 26.6 64 6.77 57 8.34 113 16.2 118 1.43 58 2.42 54 5.55 49 2.84 50
NL-TV-NCC [25]43.9 5.44 46 21.7 42 2.24 22 4.00 35 21.9 45 1.69 14 5.27 34 17.8 36 0.67 3 2.52 12 19.1 22 0.67 4 8.37 70 12.5 80 5.12 66 11.5 94 32.0 90 9.19 87 0.86 33 4.93 44 1.35 56 2.16 51 6.46 55 1.63 13
S2D-Matching [84]44.5 4.97 32 21.3 40 3.55 56 4.74 55 23.6 59 3.35 47 6.50 50 20.9 51 3.46 47 3.49 32 20.4 28 1.60 36 7.07 30 10.0 30 4.22 35 7.82 55 23.1 40 6.87 62 1.78 67 5.90 64 2.12 67 1.30 20 3.14 21 2.74 45
TV-L1-MCT [64]45.9 4.69 28 18.9 24 3.60 58 5.64 74 25.6 73 4.21 74 5.53 38 18.1 40 3.23 42 3.04 20 19.9 24 1.35 27 7.49 45 10.6 43 4.91 61 8.34 65 22.8 38 7.50 75 0.79 31 2.61 6 3.57 84 1.73 44 3.45 27 3.26 60
MDP-Flow [26]46.0 5.65 50 24.7 55 4.93 85 3.70 29 17.6 19 3.40 52 5.47 36 18.7 44 4.66 63 3.87 44 24.3 56 1.88 49 7.12 32 9.89 28 5.00 64 6.17 28 25.9 57 4.66 14 0.61 24 5.65 61 0.05 12 3.28 69 8.39 69 3.45 65
AggregFlow [97]46.2 6.17 56 23.3 50 2.58 30 7.01 84 28.0 89 5.29 85 8.46 64 24.2 60 7.66 85 3.73 38 20.2 26 1.73 40 7.25 39 10.6 43 3.52 19 4.43 5 16.4 6 4.80 17 0.75 30 5.43 56 0.25 28 1.92 47 4.46 42 4.12 71
IROF-TV [53]47.4 5.22 41 22.6 48 3.59 57 4.80 57 24.2 65 3.73 69 5.71 41 18.4 43 3.64 52 4.19 59 25.7 69 1.92 53 7.63 52 10.7 46 5.26 68 9.22 80 30.2 80 6.60 49 0.30 6 2.86 10 0.02 3 1.32 22 3.76 32 2.27 35
CombBMOF [113]49.0 6.51 61 28.6 69 2.61 31 3.98 34 18.7 28 2.29 27 5.29 35 17.4 32 2.33 27 5.12 74 26.1 74 3.28 80 6.35 16 9.81 26 3.34 16 12.0 96 28.4 72 15.1 108 3.73 84 12.8 107 0.76 42 0.98 12 3.00 19 0.09 5
OFH [38]49.7 6.38 60 25.7 58 4.69 81 3.90 32 20.6 39 2.24 26 7.85 58 24.2 60 2.27 23 4.11 57 25.1 60 1.72 39 7.44 43 10.4 40 4.69 48 8.13 62 28.9 74 8.44 83 0.44 16 4.25 31 0.12 20 2.80 60 8.82 77 2.74 45
Sparse Occlusion [54]49.9 4.99 34 21.1 36 2.79 36 4.13 38 20.1 38 3.00 39 5.94 46 19.4 47 2.15 22 3.41 30 21.8 32 1.35 27 8.17 66 12.1 74 4.74 51 7.87 59 25.6 56 6.34 46 11.4 118 17.7 120 2.71 76 1.64 40 4.70 46 1.81 21
Adaptive [20]50.0 5.12 38 22.0 43 2.34 24 4.82 59 23.2 54 3.50 58 8.67 69 24.5 65 3.56 51 4.19 59 25.3 66 1.83 47 7.40 42 10.6 43 3.63 21 5.84 24 23.2 42 3.75 8 3.25 81 8.86 83 0.89 45 2.87 62 6.69 58 3.14 59
Occlusion-TV-L1 [63]50.1 5.23 42 22.2 45 2.36 27 4.40 43 21.2 42 3.39 51 8.46 64 24.8 66 3.83 56 3.92 48 24.8 58 1.74 41 9.11 82 13.1 92 5.75 73 4.65 7 23.9 45 3.52 6 1.27 51 3.13 15 0.44 34 3.56 75 8.92 78 3.28 61
RFlow [90]50.9 5.85 53 24.8 57 4.44 78 3.18 16 17.9 21 1.88 19 7.81 57 24.4 64 2.32 26 3.25 25 23.4 49 1.55 32 7.94 56 11.6 64 4.86 57 8.23 63 28.0 71 6.64 52 1.16 47 2.13 2 1.13 52 4.10 85 9.22 84 6.81 91
ACK-Prior [27]51.8 5.49 48 24.0 53 1.81 5 2.55 6 15.7 8 0.83 1 5.07 26 17.7 35 1.52 15 2.14 6 18.1 16 0.50 1 8.64 73 11.6 64 7.10 86 14.6 107 30.7 82 11.7 97 8.46 114 11.5 99 19.5 116 3.68 79 7.25 61 2.64 44
2DHMM-SAS [92]52.1 5.14 40 21.0 32 3.79 63 5.26 69 25.2 70 3.45 55 6.97 55 20.2 48 4.18 61 4.06 55 23.3 48 2.10 63 7.18 35 10.2 35 4.92 62 8.29 64 23.7 43 7.16 69 1.26 49 5.41 55 1.63 61 1.71 43 3.75 31 2.74 45
SimpleFlow [49]52.7 5.65 50 22.4 47 4.93 85 5.47 73 24.5 68 4.28 77 6.88 54 21.0 53 3.95 57 4.74 68 25.2 62 3.02 74 7.19 36 10.1 32 4.70 49 8.34 65 23.1 40 7.16 69 1.02 40 4.61 36 0.89 45 1.29 19 3.44 25 2.47 41
S2F-IF [123]55.1 9.49 93 37.6 99 4.93 85 4.81 58 25.6 73 3.34 46 8.25 62 26.1 69 6.40 73 4.99 72 25.6 68 2.93 72 7.80 55 11.0 54 4.90 60 5.61 16 24.9 53 5.83 30 0.62 26 5.35 54 0.22 26 1.43 27 4.11 37 1.67 15
Complementary OF [21]55.2 7.27 70 30.0 74 4.31 73 3.18 16 18.9 31 1.52 10 5.91 44 20.2 48 2.31 25 4.22 62 24.8 58 2.05 59 7.50 47 10.4 40 4.99 63 12.3 99 31.7 89 8.87 84 0.61 24 2.69 9 1.72 63 3.33 70 9.22 84 4.88 83
ROF-ND [107]55.9 6.70 62 27.6 65 3.53 54 3.08 14 16.0 10 1.73 15 5.81 43 18.3 42 1.58 16 3.81 42 18.4 17 2.20 64 9.45 87 14.0 104 6.31 80 11.3 93 29.6 78 7.27 72 9.92 116 10.8 91 7.29 101 1.53 36 3.44 25 1.64 14
TCOF [69]56.6 7.04 68 26.9 62 3.54 55 4.93 60 23.7 61 3.45 55 9.94 81 27.8 77 7.40 82 3.74 39 23.7 52 1.55 32 10.0 96 14.3 105 4.40 42 4.91 8 17.0 8 5.53 22 5.08 94 9.68 87 4.19 88 1.43 27 4.44 41 1.69 17
PGM-C [120]57.9 9.47 91 37.1 96 4.81 84 5.08 63 26.1 78 3.63 62 8.75 71 27.6 75 7.02 77 5.65 82 28.1 89 3.63 87 7.99 59 11.3 59 4.88 58 5.71 20 24.5 47 5.97 33 0.31 7 3.01 13 0.02 3 2.07 49 6.50 56 2.14 31
TF+OM [100]58.1 6.03 55 23.7 52 2.78 35 4.39 42 19.9 35 3.57 59 8.73 70 23.0 57 11.2 91 3.57 35 23.2 46 1.36 29 7.98 58 11.1 56 5.89 75 8.95 77 25.3 55 7.06 67 1.68 65 11.2 96 0.20 25 3.56 75 8.35 68 4.18 72
Steered-L1 [118]58.1 4.54 23 21.2 38 2.09 16 2.13 2 13.9 4 1.31 6 4.80 20 16.5 22 1.64 18 3.87 44 25.1 60 1.60 36 8.62 72 11.5 63 7.01 83 11.1 90 28.7 73 10.4 93 12.0 122 12.3 104 34.9 124 5.90 96 9.03 82 11.6 104
FlowFields [110]59.6 9.65 95 37.6 99 5.13 88 5.09 64 25.9 76 3.72 68 8.92 73 28.3 81 7.07 79 5.45 78 26.0 72 3.82 88 7.95 57 11.2 58 5.01 65 5.75 21 26.1 60 6.01 34 0.40 14 3.29 17 0.12 20 1.92 47 5.99 53 1.89 24
DeepFlow2 [108]60.2 6.80 64 28.5 68 2.99 40 5.20 68 22.9 49 3.60 61 8.88 72 26.2 70 5.75 70 5.76 84 26.8 79 3.41 84 7.34 41 10.7 46 3.58 20 5.86 26 24.8 50 6.22 41 1.02 40 3.78 26 3.08 82 4.35 88 9.84 88 5.80 87
EPPM w/o HM [88]60.6 8.62 85 33.5 88 3.62 59 3.58 25 19.7 33 1.93 21 6.19 48 20.5 50 1.64 18 4.64 67 25.2 62 2.54 67 7.60 51 10.4 40 5.81 74 11.2 91 31.6 87 9.82 91 6.91 106 8.93 84 15.9 112 1.48 33 4.06 36 2.01 27
CPM-Flow [116]61.0 9.47 91 37.1 96 4.79 82 5.15 66 26.3 80 3.67 64 8.59 68 27.1 73 7.00 76 5.59 80 27.8 86 3.57 85 7.99 59 11.3 59 4.74 51 5.70 19 24.1 46 6.05 36 0.48 18 4.29 33 0.02 3 2.76 59 7.63 64 4.11 70
EpicFlow [102]61.9 9.44 90 37.1 96 4.80 83 5.15 66 26.4 81 3.70 66 9.58 77 30.0 84 7.07 79 5.38 77 27.8 86 3.29 81 8.01 63 11.3 59 4.88 58 5.67 18 24.6 48 6.12 38 0.32 9 3.13 15 0.02 3 3.10 67 7.52 62 4.79 80
ComplOF-FED-GPU [35]62.4 6.96 66 30.7 77 3.33 48 4.74 55 24.9 69 2.66 30 6.71 52 22.4 54 2.45 31 4.44 64 26.2 75 2.05 59 7.50 47 10.7 46 4.20 34 9.78 81 34.0 95 9.47 89 2.42 72 4.74 38 6.63 99 3.09 65 9.17 83 3.91 69
SRR-TVOF-NL [91]63.6 7.45 73 28.6 69 3.09 44 6.20 78 26.1 78 3.90 72 9.82 79 28.4 82 5.78 71 3.96 52 23.5 50 1.55 32 7.55 49 10.8 49 5.27 69 9.21 79 26.7 65 7.25 71 5.74 98 11.5 99 4.01 86 1.30 20 3.49 28 2.19 34
F-TV-L1 [15]63.9 8.70 86 31.4 82 8.47 101 7.61 88 27.3 86 5.86 86 11.0 84 28.0 78 5.73 69 5.75 83 28.7 91 3.32 83 7.28 40 10.8 49 3.72 22 6.59 32 26.4 61 4.38 11 1.26 49 5.30 52 0.44 34 3.04 64 7.76 65 2.29 38
TV-L1-improved [17]64.0 5.52 49 23.4 51 3.42 51 4.13 38 20.8 40 2.96 37 8.29 63 24.2 60 3.64 52 4.06 55 24.4 57 1.77 44 8.34 69 12.1 74 4.15 33 13.7 103 38.4 104 14.9 105 4.40 90 10.1 89 2.14 68 3.33 70 8.42 70 3.40 63
SIOF [67]64.9 5.37 44 22.6 48 2.34 24 6.11 76 28.4 90 4.30 78 12.6 90 29.2 83 14.4 92 5.52 79 27.4 83 3.00 73 8.96 81 12.6 82 6.02 76 8.72 71 27.9 69 7.93 76 0.38 11 3.48 18 0.02 3 3.09 65 7.58 63 4.85 82
Aniso. Huber-L1 [22]65.5 5.98 54 24.2 54 3.23 47 8.53 92 27.3 86 7.91 91 9.64 78 25.6 68 5.52 67 5.00 73 25.7 69 2.75 70 8.66 74 12.8 87 4.74 51 7.60 49 24.8 50 3.51 5 3.65 83 7.24 77 3.00 81 2.57 56 6.69 58 2.86 51
DPOF [18]65.6 9.01 88 34.7 90 3.68 62 6.16 77 25.4 72 4.32 79 5.55 40 17.9 38 3.36 45 3.92 48 25.3 66 2.00 57 8.14 65 11.0 54 6.05 77 10.5 86 27.9 69 8.16 78 9.33 115 6.19 66 21.0 118 1.46 31 4.57 44 0.80 9
BriefMatch [124]66.9 4.78 31 21.0 32 2.40 29 4.00 35 19.8 34 2.68 31 5.13 27 17.5 34 2.41 29 3.23 22 22.1 33 1.28 25 9.81 94 12.0 71 13.1 114 17.2 109 33.8 94 17.8 111 7.84 109 12.7 106 22.3 119 8.01 108 10.5 94 16.1 115
Classic++ [32]67.3 5.46 47 22.2 45 4.35 75 4.66 50 22.1 46 3.57 59 8.00 60 24.3 63 5.06 64 4.21 61 25.2 62 2.01 58 8.77 77 12.7 85 5.47 70 9.03 78 30.2 80 7.29 73 2.92 79 7.73 79 3.10 83 3.83 82 8.53 71 3.87 68
LocallyOriented [52]67.4 8.05 79 30.6 76 3.63 61 8.09 90 30.8 96 6.17 88 12.3 89 32.3 90 7.04 78 4.88 71 25.2 62 2.88 71 8.80 79 12.7 85 4.27 38 5.41 14 20.4 18 6.07 37 1.35 54 6.03 65 0.99 50 3.73 81 8.62 73 4.18 72
CRTflow [80]69.9 7.63 76 31.8 85 3.42 51 4.40 43 21.2 42 2.97 38 8.99 74 26.6 71 4.11 60 4.86 70 26.5 76 2.57 68 7.99 59 11.7 66 3.26 13 18.0 113 40.2 107 22.2 116 1.47 59 4.45 35 2.51 74 4.73 90 11.4 99 7.30 92
DeepFlow [86]70.0 7.55 74 29.3 72 4.67 80 6.29 80 23.7 61 4.86 81 10.0 82 28.0 78 8.76 90 6.15 91 27.3 81 3.83 89 7.49 45 10.8 49 3.72 22 6.40 30 26.8 66 6.85 61 1.12 46 2.92 12 3.94 85 7.07 101 11.2 98 12.7 106
TriFlow [95]70.1 7.87 78 30.1 75 3.19 46 7.12 85 24.4 67 7.15 90 13.9 93 31.4 87 20.0 97 3.50 33 22.4 35 1.77 44 8.70 75 11.7 66 7.03 85 7.51 47 21.9 30 6.63 51 28.6 125 14.7 115 78.3 127 2.16 51 5.57 50 2.14 31
Rannacher [23]71.8 6.99 67 27.1 63 5.36 91 5.27 70 24.3 66 4.22 75 9.51 76 27.1 73 5.54 68 4.76 69 25.7 69 2.58 69 8.80 79 12.9 89 4.82 56 11.0 88 35.7 98 9.36 88 2.33 71 4.76 40 2.39 73 2.82 61 8.01 66 3.13 58
Brox et al. [5]72.1 8.32 82 32.6 86 6.95 96 6.23 79 26.9 85 5.23 83 9.13 75 27.6 75 6.55 75 5.85 86 28.2 90 3.26 78 10.2 98 12.9 89 11.0 109 5.43 15 29.3 77 4.79 16 0.86 33 4.00 28 0.12 20 4.32 86 10.2 91 4.54 78
Bartels [41]72.6 6.83 65 26.2 60 5.19 89 3.93 33 17.4 18 3.30 44 6.63 51 22.6 55 3.25 43 4.45 65 23.9 53 2.48 66 9.12 83 12.1 74 8.25 94 10.6 87 31.1 83 12.3 100 5.74 98 10.4 90 18.9 115 5.34 92 9.52 86 8.47 98
Local-TV-L1 [65]72.6 9.60 94 30.8 78 7.89 100 12.7 98 30.2 95 13.3 98 15.9 99 32.3 90 17.3 94 6.19 92 28.0 88 3.84 90 7.55 49 10.9 53 4.22 35 7.48 44 26.4 61 6.02 35 0.28 3 1.87 1 0.15 23 9.10 110 10.8 96 20.5 116
SuperFlow [81]73.2 7.15 69 27.4 64 3.52 53 10.4 95 27.8 88 11.2 95 14.5 97 31.5 89 22.4 99 5.93 87 31.6 97 3.23 77 8.77 77 11.9 68 8.59 98 5.61 16 25.9 57 3.72 7 3.76 86 11.1 95 0.37 33 3.59 77 8.96 80 3.01 54
Dynamic MRF [7]73.4 7.74 77 31.6 84 4.44 78 4.12 37 23.6 59 2.47 29 8.49 66 28.0 78 2.83 39 4.25 63 27.4 83 2.41 65 8.61 71 12.0 71 6.08 78 14.5 106 43.2 111 14.9 105 0.64 27 2.35 3 4.51 91 9.85 113 15.6 115 15.3 113
CBF [12]77.5 6.32 57 26.2 60 3.35 49 11.1 96 25.6 73 13.7 99 8.51 67 24.1 59 7.12 81 5.12 74 26.0 72 3.04 76 10.3 99 13.6 99 9.59 105 7.85 58 26.4 61 4.25 9 11.8 119 13.8 110 14.2 111 3.54 74 8.06 67 5.32 85
CLG-TV [48]78.4 6.33 58 24.7 55 4.13 69 9.08 94 26.6 82 9.31 94 9.85 80 26.8 72 5.82 72 5.30 76 26.5 76 3.03 75 10.4 101 14.6 109 7.57 89 7.95 60 31.1 83 6.51 48 5.92 100 11.4 97 4.36 89 3.41 72 8.81 76 3.06 55
TriangleFlow [30]78.9 7.35 71 28.2 67 4.31 73 5.35 72 25.2 70 3.36 48 8.00 60 24.8 66 2.70 35 3.90 47 24.1 55 1.97 56 12.9 115 17.8 121 10.7 107 13.1 101 32.3 92 13.9 103 4.71 92 16.1 117 4.04 87 3.65 78 8.73 75 5.69 86
DF-Auto [115]80.0 9.74 96 34.1 89 4.36 76 14.1 101 31.9 99 15.4 101 15.6 98 33.1 96 23.6 101 5.94 88 27.3 81 3.59 86 10.4 101 14.8 112 6.97 82 3.80 3 21.1 24 2.46 3 5.25 97 11.4 97 0.49 36 4.33 87 10.4 92 4.33 74
p-harmonic [29]80.4 8.47 83 36.3 94 7.17 97 5.27 70 24.1 64 4.39 80 11.2 86 31.4 87 8.13 89 7.18 97 32.4 98 5.24 98 8.04 64 11.1 56 6.89 81 9.82 82 36.4 100 10.6 94 2.61 74 5.51 58 0.54 38 4.07 84 9.01 81 4.34 75
CNN-flow-warp+ref [117]81.1 9.81 98 35.7 92 7.67 99 8.14 91 26.0 77 8.55 92 14.3 95 35.8 99 15.7 93 6.69 94 30.3 94 4.31 94 9.17 85 12.2 77 8.71 100 7.03 38 29.6 78 5.55 25 0.69 28 3.77 25 2.00 66 7.79 106 12.1 103 8.13 97
FlowNet2 [122]82.6 21.7 112 43.8 106 13.5 105 24.6 113 42.3 112 27.3 114 19.8 102 40.5 102 29.9 107 8.21 101 23.6 51 5.39 100 9.85 95 12.6 82 8.54 96 8.76 72 28.9 74 5.89 32 2.77 76 15.5 116 0.81 44 1.28 18 4.68 45 0.26 7
FlowNetS+ft+v [112]83.9 7.57 75 29.4 73 3.96 66 7.50 87 26.6 82 6.48 89 14.3 95 32.7 94 17.5 95 7.55 98 31.3 95 5.28 99 10.5 103 14.7 110 7.49 88 6.75 35 27.8 68 6.97 65 4.01 88 8.84 82 6.77 100 3.52 73 9.71 87 3.61 66
SegOF [10]84.6 12.6 101 34.9 91 7.20 98 21.3 109 36.9 107 25.3 112 21.6 105 40.5 102 31.8 111 14.1 109 37.7 106 10.8 104 10.3 99 12.5 80 12.6 113 10.2 84 40.2 107 11.2 96 0.29 4 2.91 11 0.07 15 2.90 63 8.68 74 2.07 30
LDOF [28]85.3 8.22 80 31.4 82 4.08 67 7.64 89 29.4 92 5.87 87 10.7 83 30.3 85 7.99 88 7.80 99 36.8 104 4.86 97 9.14 84 12.4 79 8.24 93 8.58 68 32.0 90 8.38 81 1.75 66 5.26 51 5.02 92 5.52 94 12.9 107 6.04 89
Fusion [6]85.5 8.51 84 37.6 99 6.69 95 3.62 27 20.0 37 3.08 40 6.82 53 22.6 55 6.47 74 5.78 85 31.3 95 4.29 93 11.2 111 14.7 110 10.6 106 14.0 104 35.2 96 15.0 107 7.88 110 14.3 113 2.22 69 5.35 93 11.0 97 8.56 99
Learning Flow [11]87.7 6.74 63 28.1 66 3.03 42 6.37 81 28.7 91 5.02 82 11.8 87 32.6 93 7.93 87 6.87 96 33.2 101 4.32 95 12.5 114 17.4 119 7.78 90 9.98 83 35.2 96 8.41 82 2.66 75 10.9 93 2.24 70 6.76 100 13.7 109 6.41 90
StereoFlow [44]88.7 58.0 127 76.4 127 63.7 124 51.8 126 66.9 127 48.3 122 51.0 127 73.0 127 41.6 121 63.5 127 83.4 127 56.7 125 13.3 117 13.7 100 19.1 120 3.63 2 20.4 18 2.73 4 0.26 1 2.49 5 0.05 12 4.06 83 8.57 72 5.81 88
Ad-TV-NDC [36]89.5 21.2 111 36.8 95 34.1 119 25.9 115 38.5 109 29.9 115 23.5 109 41.0 104 27.1 102 13.3 106 32.4 98 13.3 109 8.75 76 13.2 93 3.82 27 7.43 43 25.1 54 6.92 64 1.50 61 4.84 42 0.34 31 17.1 122 15.9 119 37.2 125
Second-order prior [8]89.9 7.35 71 31.2 81 4.16 70 6.80 83 29.5 93 5.27 84 11.8 87 33.3 97 7.78 86 6.05 89 27.2 80 3.90 91 9.67 92 13.8 102 5.74 72 14.0 104 41.8 110 11.7 97 6.86 103 9.72 88 7.61 104 4.72 89 10.1 90 7.78 95
HBpMotionGpu [43]90.9 11.7 99 32.8 87 6.34 94 18.9 106 35.4 105 22.0 109 22.3 106 42.7 107 31.1 110 5.62 81 26.7 78 3.31 82 9.47 88 13.0 91 8.55 97 8.68 69 31.2 85 5.58 26 6.88 105 11.9 101 0.64 40 7.67 105 11.4 99 15.2 111
StereoOF-V1MT [119]92.9 9.29 89 44.8 108 4.17 71 7.22 86 34.5 102 3.68 65 13.7 92 42.6 106 3.80 55 6.06 90 38.5 108 3.27 79 11.0 108 15.1 114 9.55 104 15.1 108 49.9 117 14.0 104 1.08 44 5.51 58 5.44 94 9.33 112 15.5 114 9.73 102
SPSA-learn [13]93.0 15.7 106 48.8 112 16.5 107 16.6 103 35.0 104 17.5 104 21.4 104 42.3 105 29.7 106 12.6 104 37.4 105 12.3 107 9.64 90 12.8 87 9.16 101 11.0 88 37.9 103 12.2 99 0.98 38 3.88 27 0.05 12 8.38 109 11.6 101 15.2 111
Shiralkar [42]93.1 9.76 97 46.6 110 4.40 77 6.53 82 31.3 98 4.04 73 12.7 91 37.5 100 5.34 66 6.47 93 32.9 100 4.34 96 8.33 68 11.9 68 5.58 71 17.4 112 43.3 113 15.5 109 6.82 102 8.77 81 14.0 110 7.36 103 15.7 118 7.83 96
Filter Flow [19]95.1 14.6 103 38.2 103 8.96 103 12.4 97 34.6 103 11.3 96 20.2 103 38.3 101 30.1 108 19.2 111 43.4 112 18.6 112 10.0 96 13.4 95 9.43 103 10.3 85 31.4 86 9.08 86 8.21 112 19.6 121 0.79 43 3.72 80 6.85 60 3.41 64
Modified CLG [34]95.3 15.7 106 43.7 105 12.2 104 19.1 107 33.3 101 23.7 110 25.1 110 47.4 110 35.6 116 13.2 105 35.5 103 11.1 105 10.7 105 14.4 107 9.36 102 7.26 40 35.8 99 6.19 40 1.66 64 5.21 50 6.43 97 5.94 97 13.8 111 7.73 94
GraphCuts [14]96.6 12.6 101 36.1 93 5.46 92 14.7 102 39.4 111 12.5 97 17.8 101 35.6 98 29.1 105 6.86 95 33.7 102 4.15 92 9.33 86 12.6 82 8.69 99 23.0 118 31.6 87 15.5 109 3.52 82 7.38 78 11.7 108 5.33 91 9.87 89 8.75 100
2D-CLG [1]96.7 24.4 114 51.8 115 19.4 111 27.4 116 38.7 110 33.8 118 34.6 117 57.7 116 42.2 123 33.4 121 57.1 122 32.9 120 9.64 90 12.2 77 11.0 109 11.2 91 40.2 107 12.8 101 0.31 7 2.62 7 0.25 28 6.33 98 13.7 109 7.33 93
IAOF2 [51]96.8 8.72 87 30.9 79 5.32 90 13.9 100 31.1 97 15.4 101 14.1 94 33.0 95 18.2 96 30.8 119 42.2 111 36.4 121 9.74 93 13.8 102 6.09 79 12.0 96 33.4 93 7.96 77 7.92 111 13.9 111 7.49 103 5.69 95 10.6 95 4.51 77
BlockOverlap [61]99.5 12.3 100 29.2 71 8.49 102 13.7 99 29.6 94 15.3 100 16.2 100 32.5 92 20.0 97 8.87 102 27.4 83 7.62 102 10.9 107 13.4 95 12.5 112 13.3 102 29.1 76 10.3 92 11.8 119 14.4 114 23.8 120 10.6 114 8.92 78 24.8 118
IAOF [50]101.1 14.7 104 37.8 102 14.8 106 17.3 105 33.2 100 18.7 105 22.7 107 44.3 108 23.3 100 20.9 113 38.7 109 24.5 116 9.60 89 13.3 94 8.28 95 13.0 100 38.8 105 7.37 74 4.20 89 7.90 80 2.59 75 14.5 119 13.4 108 32.0 123
Black & Anandan [4]102.1 15.1 105 45.4 109 18.1 110 16.6 103 36.3 106 16.9 103 23.3 108 44.9 109 27.8 103 13.5 107 38.1 107 13.1 108 11.1 110 15.7 115 7.97 91 11.6 95 39.6 106 11.0 95 5.17 96 9.06 85 2.27 71 7.28 102 12.3 104 10.2 103
2bit-BM-tele [98]103.0 20.3 110 39.1 104 26.1 115 8.84 93 26.8 84 9.29 93 11.1 85 30.7 86 7.60 84 8.06 100 29.9 93 5.91 101 11.0 108 13.7 100 11.8 111 18.0 113 37.1 101 19.8 115 17.1 123 20.4 123 30.8 123 6.54 99 11.7 102 11.8 105
UnFlow [129]103.0 45.7 124 58.8 116 25.7 114 28.2 117 44.0 113 31.2 117 38.6 123 68.3 125 37.0 118 19.4 112 46.0 113 16.6 111 13.8 119 14.9 113 18.2 119 20.9 116 49.2 116 23.5 117 2.86 77 6.76 72 0.22 26 3.12 68 10.4 92 2.27 35
GroupFlow [9]103.8 22.9 113 47.1 111 26.7 117 28.4 118 50.0 119 30.8 116 25.4 111 52.4 112 30.6 109 9.32 103 29.6 92 8.14 103 10.7 105 13.4 95 7.16 87 23.0 118 46.3 114 27.8 121 1.56 62 5.72 63 2.76 77 8.00 107 12.5 105 15.3 113
Nguyen [33]104.0 20.0 109 44.7 107 17.4 109 39.5 122 37.5 108 52.5 123 34.0 116 56.0 115 38.8 119 35.6 122 47.9 115 41.1 122 12.1 112 14.3 105 16.5 117 12.0 96 37.8 102 13.8 102 1.37 56 4.27 32 0.71 41 11.6 117 14.5 113 20.8 117
SILK [79]108.8 26.9 115 51.7 114 36.6 121 22.3 111 45.5 115 24.5 111 28.8 112 54.7 114 34.2 112 18.4 110 41.6 110 15.8 110 13.1 116 16.5 116 16.1 116 19.1 115 47.8 115 19.3 114 2.87 78 4.22 30 6.53 98 15.9 120 19.1 120 25.8 119
Heeger++ [104]109.8 42.8 122 66.4 126 26.2 116 25.0 114 60.7 126 19.8 108 38.4 122 66.9 122 28.0 104 23.5 115 49.3 116 19.7 113 10.6 104 13.4 95 8.12 92 40.8 125 67.2 127 45.1 125 2.04 70 10.9 93 1.70 62 11.2 115 15.6 115 12.8 107
Horn & Schunck [3]111.2 19.9 108 61.0 120 23.3 112 19.4 108 44.3 114 19.1 106 29.5 113 58.8 118 34.9 114 21.0 114 49.9 117 21.2 114 12.3 113 16.5 116 10.8 108 17.3 111 50.6 119 18.0 112 7.23 108 11.9 101 2.34 72 13.4 118 22.2 122 14.9 110
Periodicity [78]114.8 30.6 118 48.8 112 16.7 108 24.1 112 49.8 117 26.2 113 39.1 124 54.5 113 39.5 120 13.6 108 47.1 114 12.0 106 37.5 127 48.2 127 33.6 126 38.5 124 66.9 126 36.0 124 2.02 69 10.8 91 8.18 105 20.8 123 35.9 126 30.1 121
TI-DOFE [24]115.5 44.7 123 66.3 125 66.5 126 44.2 124 50.5 120 54.8 125 43.5 126 72.0 126 44.7 125 48.6 124 63.3 123 54.0 124 13.6 118 17.7 120 15.1 115 17.2 109 50.3 118 19.0 113 3.07 80 5.50 57 2.93 80 21.5 124 24.7 124 33.9 124
FFV1MT [106]115.5 39.5 120 59.3 117 25.4 113 21.8 110 56.3 125 19.1 106 38.0 121 67.0 123 34.9 114 24.3 116 55.7 121 22.2 115 17.7 123 18.9 123 25.5 124 41.8 126 66.3 125 45.5 126 3.73 84 12.9 108 6.35 96 11.2 115 15.6 115 12.8 107
SLK [47]117.0 28.9 117 63.3 123 36.4 120 42.3 123 54.0 124 52.8 124 36.6 119 67.7 124 42.5 124 51.4 125 54.3 120 60.0 126 14.5 121 16.7 118 20.8 122 21.5 117 53.4 123 24.1 118 3.92 87 6.27 68 5.91 95 21.7 125 23.7 123 31.5 122
Adaptive flow [45]119.8 49.6 125 62.1 121 66.8 127 37.4 120 46.5 116 43.1 120 34.9 118 58.6 117 41.7 122 27.3 118 53.5 119 28.7 117 16.1 122 18.2 122 17.6 118 25.3 122 52.9 121 25.1 120 45.4 126 38.1 126 74.4 125 9.25 111 14.4 112 13.9 109
PGAM+LK [55]120.1 35.2 119 65.7 124 44.1 123 31.5 119 51.1 121 36.1 119 30.9 114 60.0 120 36.8 117 33.0 120 72.3 126 32.7 119 13.8 119 14.4 107 22.6 123 24.6 120 53.2 122 24.6 119 27.1 124 32.6 125 26.2 121 17.0 121 20.4 121 28.4 120
FOLKI [16]120.3 27.4 116 59.9 118 40.6 122 37.4 120 51.5 122 46.6 121 32.4 115 61.6 121 34.2 112 26.3 117 50.6 118 30.2 118 18.2 125 19.7 124 26.3 125 24.6 120 56.6 124 28.8 122 10.3 117 13.9 111 26.7 122 27.1 126 26.9 125 45.3 126
HCIC-L [99]120.8 51.6 126 60.9 119 30.9 118 58.4 127 53.4 123 73.0 127 39.8 125 50.8 111 52.8 127 63.4 126 71.0 125 69.9 127 18.1 124 20.6 125 20.5 121 29.9 123 43.2 111 34.1 123 73.0 127 62.0 127 76.4 126 7.45 104 12.5 105 8.87 101
Pyramid LK [2]123.5 41.0 121 62.5 122 66.4 125 47.2 125 49.9 118 59.7 126 37.5 120 59.5 119 45.5 126 43.8 123 65.1 124 49.5 123 36.5 126 43.8 126 42.6 127 43.3 127 52.8 120 45.9 127 11.8 119 20.2 122 20.7 117 40.0 127 46.5 127 59.5 127
AdaConv-v1 [126]128.0 99.3 128 97.8 128 99.8 128 99.9 128 100.0 128 99.8 128 99.9 128 99.9 128 99.9 128 99.5 128 99.9 128 99.9 128 99.9 128 99.9 128 99.9 128 99.1 128 98.9 128 99.7 128 98.5 128 93.0 128 100.0 128 99.9 128 99.9 128 99.9 128
SepConv-v1 [127]128.0 99.3 128 97.8 128 99.8 128 99.9 128 100.0 128 99.8 128 99.9 128 99.9 128 99.9 128 99.5 128 99.9 128 99.9 128 99.9 128 99.9 128 99.9 128 99.1 128 98.9 128 99.7 128 98.5 128 93.0 128 100.0 128 99.9 128 99.9 128 99.9 128
Move the mouse over the numbers in the table to see the corresponding images. Click to compare with the ground truth.

References

Methodtime*framescolor Reference and notes
[1] 2D-CLG 844 2 gray The 2D-CLG method by Bruhn et al. as implemented by Stefan Roth. [A. Bruhn, J. Weickert, and C. Schnörr. Lucas/Kanade meets Horn/Schunck: combining local and global optic flow methods. IJCV 63(3), 2005.] Parameters were set to match the published performance on Yosemite sequence, which may not be optimal for other sequences.
[2] Pyramid LK 12 2 color A modification of Bouguet's pyramidal implementation of Lucas-Kanade.
[3] Horn & Schunck 49 2 gray A modern Matlab implementation of the Horn & Schunck method by Deqing Sun. Parameters set to optimize AAE on all training data.
[4] Black & Anandan 328 2 gray A modern Matlab implementation of the Black & Anandan method by Deqing Sun.
[5] Brox et al. 18 2 color T. Brox, A. Bruhn, N. Papenberg, and J. Weickert. High accuracy optical flow estimation based on a theory for warping. ECCV 2004. (Improved using separate robust functions as proposed in A. Bruhn and J. Weickert, Towards ultimate motion estimation, ICCV 2005; improved by training on the training set.)
[6] Fusion 2,666 2 color V. Lempitsky, S. Roth, and C. Rother. Discrete-continuous optimization for optical flow estimation. CVPR 2008.
[7] Dynamic MRF 366 2 gray B. Glocker, N. Paragios, N. Komodakis, G. Tziritas, and N. Navab. Optical flow estimation with uncertainties through dynamic MRFs. CVPR 2008. (Method improved since publication.)
[8] Second-order prior 14 2 gray W. Trobin, T. Pock, D. Cremers, and H. Bischof. An unbiased second-order prior for high-accuracy motion estimation. DAGM 2008. (Method improved since publication; for details see W. Trobin, Ph.D. thesis, 2009.)
[9] GroupFlow 600 2 gray X. Ren. Local Grouping for Optical Flow. CVPR 2008.
[10] SegOF 60 2 color L. Xu, J. Chen, and J. Jia. Segmentation based variational model for accurate optical flow estimation. ECCV 2008. Code available.
[11] Learning Flow 825 2 gray D. Sun, S. Roth, J.P. Lewis, and M. Black. Learning optical flow (SRF-LFC). ECCV 2008.
[12] CBF 69 2 color W. Trobin, T. Pock, D. Cremers, and H. Bischof. Continuous energy minimization via repeated binary fusion. ECCV 2008. (Method improved since publication; for details see W. Trobin, Ph.D. thesis, 2009.)
[13] SPSA-learn 200 2 color Y. Li and D. Huttenlocher. Learning for optical flow using stochastic optimization. ECCV 2008.
[14] GraphCuts 1,200 2 color T. Cooke. Two applications of graph-cuts to image processing. DICTA 2008.
[15] F-TV-L1 8 2 gray A. Wedel, T. Pock, J. Braun, U. Franke, and D. Cremers. Duality TV-L1 flow with fundamental matrix prior. IVCNZ 2008.
[16] FOLKI 1.4 2 gray G. Le Besnerais and F. Champagnat. Dense optical flow by iterative local window registration. ICIP 2005.
[17] TV-L1-improved 2.9 2 gray A. Wedel, T. Pock, C. Zach, H. Bischof, and D. Cremers. An improved algorithm for TV-L1 optical flow computation. Proceedings of the Dagstuhl Visual Motion Analysis Workshop 2008. Code at GPU4Vision.
[18] DPOF 287 2 color C. Lei and Y.-H. Yang. Optical flow estimation on coarse-to-fine region-trees using discrete optimization. ICCV 2009. (Method improved since publication.)
[19] Filter Flow 34,000 2 color S. Seitz and S. Baker. Filter flow. ICCV 2009.
[20] Adaptive 9.2 2 gray A. Wedel, D. Cremers, T. Pock, and H. Bischof. Structure- and motion-adaptive regularization for high accuracy optic flow. ICCV 2009.
[21] Complementary OF 44 2 color H. Zimmer, A. Bruhn, J. Weickert, L. Valgaerts, A. Salgado, B. Rosenhahn, and H.-P. Seidel. Complementary optic flow. EMMCVPR 2009.
[22] Aniso. Huber-L1 2 2 gray M. Werlberger, W. Trobin, T. Pock, A. Wedel, D. Cremers, and H. Bischof. Anisotropic Huber-L1 optical flow. BMVC 2009. Code at GPU4Vision.
[23] Rannacher 0.12 2 gray J. Rannacher. Realtime 3D motion estimation on graphics hardware. Bachelor thesis, Heidelberg University, 2009.
[24] TI-DOFE 260 2 gray C. Cassisa, S. Simoens, and V. Prinet. Two-frame optical flow formulation in an unwarped multiresolution scheme. CIARP 2009.
[25] NL-TV-NCC 20 2 color M. Werlberger, T. Pock, and H. Bischof. Motion estimation with non-local total variation regularization. CVPR 2010.
[26] MDP-Flow 188 2 color L. Xu, J. Jia, and Y. Matsushita. Motion detail preserving optical flow estimation. CVPR 2010.
[27] ACK-Prior 5872 2 color K. Lee, D. Kwon, I. Yun, and S. Lee. Optical flow estimation with adaptive convolution kernel prior on discrete framework. CVPR 2010.
[28] LDOF 122 2 color T. Brox and J. Malik. Large displacement optical flow: descriptor matching in variational motion estimation. PAMI 33(3):500-513, 2011.
[29] p-harmonic 565 2 gray J. Gai and R. Stevenson. Optical flow estimation with p-harmonic regularization. ICIP 2010.
[30] TriangleFlow 4200 2 gray B. Glocker, H. Heibel, N. Navab, P. Kohli, and C. Rother. TriangleFlow: Optical flow with triangulation-based higher-order likelihoods. ECCV 2010.
[31] Classic+NL 972 2 color D. Sun, S. Roth, and M. Black. Secrets of optical flow estimation and their principles. CVPR 2010. Matlab code.
[32] Classic++ 486 2 gray A modern implementation of the classical formulation descended from Horn & Schunck and Black & Anandan; see D. Sun, S. Roth, and M. Black, Secrets of optical flow estimation and their principles, CVPR 2010.
[33] Nguyen 33 2 gray D. Nguyen. Tuning optical flow estimation with image-driven functions. ICRA 2011.
[34] Modified CLG 133 2 gray R. Fezzani, F. Champagnat, and G. Le Besnerais. Combined local global method for optic flow computation. EUSIPCO 2010.
[35] ComplOF-FED-GPU 0.97 2 color P. Gwosdek, H. Zimmer, S. Grewenig, A. Bruhn, and J. Weickert. A highly efficient GPU implementation for variational optic flow based on the Euler-Lagrange framework. CVGPU Workshop 2010.
[36] Ad-TV-NDC 35 2 gray M. Nawaz. Motion estimation with adaptive regularization and neighborhood dependent constraint. DICTA 2010.
[37] Layers++ 18206 2 color D. Sun, E. Sudderth, and M. Black. Layered image motion with explicit occlusions, temporal consistency, and depth ordering. NIPS 2010.
[38] OFH 620 3 color H. Zimmer, A. Bruhn, J. Weickert. Optic flow in harmony. IJCV 93(3) 2011.
[39] LSM 1615 2 color K. Jia, X. Wang, and X. Tang. Optical flow estimation using learned sparse model. ICCV 2011.
[40] CostFilter 55 2 color C. Rhemann, A. Hosni, M. Bleyer, C. Rother, and M. Gelautz. Fast cost-volume filtering for visual correspondence and beyond. CVPR 2011.
[41] Bartels 0.15 2 gray C. Bartels and G. de Haan. Smoothness constraints in recursive search motion estimation for picture rate conversion. IEEE TCSVT 2010. Version improved since publication: mapped on GPU.
[42] Shiralkar 600 2 gray M. Shiralkar and R. Schalkoff. A self organization-based optical flow estimator with GPU implementation. MVA 23(6):1229-1242.
[43] HBpMotionGpu 1000 5 gray S. Grauer-Gray and C. Kambhamettu. Hierarchical belief propagation to reduce search space using CUDA for stereo and motion estimation. WACV 2009. (Method improved since publication.)
[44] StereoFlow 7200 2 color G. Rosman, S. Shem-Tov, D. Bitton, T. Nir, G. Adiv, R. Kimmel, A. Feuer, and A. Bruckstein. Over-parameterized optical flow using a stereoscopic constraint. SSVM 2011:761-772.
[45] Adaptive flow 121 2 gray T. Arici. Energy minimization based motion estimation using adaptive smoothness priors. Submitted to IEEE TIP 2011.
[46] TC-Flow 2500 5 color S. Volz, A. Bruhn, L. Valgaerts, and H. Zimmer. Modeling temporal coherence for optical flow. ICCV 2011.
[47] SLK 300 2 gray T. Corpetti and E. Mémin. Stochastic uncertainty models for the luminance consistency assumption. IEEE TIP 2011.
[48] CLG-TV 29 2 gray M. Drulea. Total variation regularization of local-global optical flow. ITSC 2011. Matlab code.
[49] SimpleFlow 1.7 2 color M. Tao, J. Bai, P. Kohli, S. Paris. SimpleFlow: a non-iterative, sublinear optical flow algorithm. EUROGRAPHICS 2012.
[50] IAOF 57 2 gray D. Nguyen. Improving motion estimation using image-driven functions and hybrid scheme. PSIVT 2011.
[51] IAOF2 56 2 gray D. Nguyen. Enhancing the sharpness of flow field using image-driven functions with occlusion-aware filter. Submitted to TIP 2011.
[52] LocallyOriented 9541 2 gray Y.Niu, A. Dick, and M. Brooks. Locally oriented optical flow computation. To appear in TIP 2012.
[53] IROF-TV 261 2 color H. Rashwan, D. Puig, and M. Garcia. On improving the robustness of differential optical flow. ICCV 2011 Artemis workshop.
[54] Sparse Occlusion 2312 2 color A. Ayvaci, M. Raptis, and S. Soatto. Sparse occlusion detection with optical flow. Submitted to IJCV 2011.
[55] PGAM+LK 0.37 2 gray A. Alba, E. Arce-Santana, and M. Rivera. Optical flow estimation with prior models obtained from phase correlation. ISVC 2010.
[56] Sparse-NonSparse 713 2 color L. Chen, J. Wang, and Y. Wu. Decomposing and regularizing sparse/non-sparse components for motion field estimation. Submitted to PAMI 2013.
[57] nLayers 36150 4 color D. Sun, E. Sudderth, and M. Black. Layered segmentation and optical flow estimation over time. CVPR 2012.
[58] IROF++ 187 2 color H. Rashwan, D. Puig, and M. Garcia. Variational optical flow estimation based on stick tensor voting. IEEE TIP 2013.
[59] COFM 600 3 color M. Mozerov. Constrained optical flow estimation as a matching problem. IEEE TIP 2013.
[60] Efficient-NL 400 2 color P. Krähenbühl and V. Koltun. Efficient nonlocal regularization for optical flow. ECCV 2012.
[61] BlockOverlap 2 2 gray M. Santoro, G. AlRegib, and Y. Altunbasak. Motion estimation using block overlap minimization. Submitted to MMSP 2012.
[62] Ramp 1200 2 color A. Singh and N. Ahuja. Exploiting ramp structures for improving optical flow estimation. ICPR 2012.
[63] Occlusion-TV-L1 538 3 gray C. Ballester, L. Garrido, V. Lazcano, and V. Caselles. A TV-L1 optical flow method with occlusion detection. DAGM-OAGM 2012.
[64] TV-L1-MCT 90 2 color M. Mohamed and B. Mertsching. TV-L1 optical flow estimation with image details recovering based on modified census transform. ISVC 2012.
[65] Local-TV-L1 500 2 gray L. Raket. Local smoothness for global optical flow. ICIP 2012.
[66] ALD-Flow 61 2 color M. Stoll, A. Bruhn, and S. Volz. Adaptive integration of feature matches into variational optic flow methods. ACCV 2012.
[67] SIOF 234 2 color L. Xu, Z. Dai, and J. Jia. Scale invariant optical flow. ECCV 2012.
[68] MDP-Flow2 342 2 color L. Xu, J. Jia, and Y. Matsushita. Motion detail preserving optical flow estimation. PAMI 34(9):1744-1757, 2012. Code available.
[69] TCOF 1421 all gray J. Sanchez, A. Salgado, and N. Monzon. Optical flow estimation with consistent spatio-temporal coherence models. VISAPP 2013.
[70] LME 476 2 color W. Li, D. Cosker, M. Brown, and R. Tang. Optical flow estimation using Laplacian mesh energy. CVPR 2013.
[71] NN-field 362 2 color L. Chen, H. Jin, Z. Lin, S. Cohen, and Y. Wu. Large displacement optical flow from nearest neighbor fields. CVPR 2013.
[72] FESL 3310 2 color W. Dong, G. Shi, X. Hu, and Y. Ma. Nonlocal sparse and low-rank regularization for optical flow estimation. Submitted to IEEE TIP 2013.
[73] PMF 35 2 color J. Lu, H. Yang, D. Min, and M. Do. PatchMatch filter: efficient edge-aware filtering meets randomized search for fast correspondence field estimation. CVPR 2013.
[74] FC-2Layers-FF 2662 4 color D. Sun, J. Wulff, E. Sudderth, H. Pfister, and M. Black. A fully-connected layered model of foreground and background flow. CVPR 2013.
[75] Correlation Flow 290 2 color M. Drulea and S. Nedevschi. Motion estimation using the correlation transform. TIP 2013. Matlab code.
[76] TC/T-Flow 341 5 color M. Stoll, S. Volz, and A. Bruhn. Joint trilateral filtering for multiframe optical flow. ICIP 2013.
[77] OFLAF 1530 2 color T. Kim, H. Lee, and K. Lee. Optical flow via locally adaptive fusion of complementary data costs. ICCV 2013.
[78] Periodicity 8000 4 color G. Khachaturov, S. Gonzalez-Brambila, and J. Gonzalez-Trejo. Periodicity-based computation of optical flow. Submitted to Computacion y Sistemas (CyS) 2013.
[79] SILK 572 2 gray P. Zille, C. Xu, T. Corpetti, L. Shao. Observation models based on scale interactions for optical flow estimation. Submitted to IEEE TIP.
[80] CRTflow 13 3 color O. Demetz, D. Hafner, and J. Weickert. The complete rank transform: a tool for accurate and morphologically invariant matching of structures. BMVC 2013.
[81] SuperFlow 178 2 color Anonymous. Superpixel based optical flow estimation. ICCV 2013 submission 507.
[82] Aniso-Texture 300 2 color Anonymous. Texture information-based optical flow estimation using an incremental multi-resolution approach. ITC-CSCC 2013 submission 267.
[83] Classic+CPF 640 2 gray Z. Tu, R. Veltkamp, and N. van der Aa. A combined post-filtering method to improve accuracy of variational optical flow estimation. Submitted to Pattern Recognition 2013.
[84] S2D-Matching 1200 2 color Anonymous. Locally affine sparse-to-dense matching for motion and occlusion estimation. ICCV 2013 submission 1479.
[85] AGIF+OF 438 2 gray Z. Tu, R. Poppe, and R. Veltkamp. Adaptive guided image filter to warped interpolation image for variational optical flow computation. Submitted to Signal Processing 2015.
[86] DeepFlow 13 2 color P. Weinzaepfel, J. Revaud, Z. Harchaoui, and C. Schmid. DeepFlow: large displacement optical flow with deep matching. ICCV 2013.
[87] NNF-Local 673 2 color Z. Chen, H. Jin, Z. Lin, S. Cohen, and Y. Wu. Large displacement optical flow with nearest neighbor field. Submitted to PAMI 2014.
[88] EPPM w/o HM 2.5 2 color L. Bao, Q. Yang, and H. Jin. Fast edge-preserving PatchMatch for large displacement optical flow. CVPR 2014.
[89] MLDP_OF 165 2 gray M. Mohamed, H. Rashwan, B. Mertsching, M. Garcia, and D. Puig. Illumination-robust optical flow approach using local directional pattern. IEEE TCSVT 24(9):1499-1508, 2014.
[90] RFlow 20 2 gray S. Ali, C. Daul, and W. Blondel. Robust and accurate optical flow estimation for weak texture and varying illumination condition: Application to cystoscopy. IPTA 2014.
[91] SRR-TVOF-NL 32 all color P. Pohl, M. Sirotenko, E. Tolstaya, and V. Bucha. Edge preserving motion estimation with occlusions correction for assisted 2D to 3D conversion. IS&T/SPIE Electronic Imaging 2014.
[92] 2DHMM-SAS 157 2 color M.-C. Shih, R. Shenoy, and K. Rose. A two-dimensional hidden Markov model with spatially-adaptive states with application of optical flow. ICIP 2014 submission.
[93] WLIF-Flow 700 2 color Z. Tu, R. Veltkamp, N. van der Aa, and C. Van Gemeren. Weighted local intensity fusion method for variational optical flow estimation. Submitted to TIP 2014.
[94] FMOF 215 2 color N. Jith, A. Ramakanth, and V. Babu. Optical flow estimation using approximate nearest neighbor field fusion. ICASSP 2014.
[95] TriFlow 150 2 color TriFlow. Optical flow with geometric occlusion estimation and fusion of multiple frames. ECCV 2014 submission 914.
[96] ComponentFusion 6.5 2 color Anonymous. Fast optical flow by component fusion. ECCV 2014 submission 941.
[97] AggregFlow 1642 2 color D. Fortun, P. Bouthemy, and C. Kervrann. Aggregation of local parametric candidates and exemplar-based occlusion handling for optical flow. Preprint arXiv:1407.5759.
[98] 2bit-BM-tele 124 2 gray R. Xu and D. Taubman. Robust dense block-based motion estimation using a two-bit transform on a Laplacian pyramid. ICIP 2013.
[99] HCIC-L 330 2 color Anonymous. Globally-optimal image correspondence using a hierarchical graphical model. NIPS 2014 submission 114.
[100] TF+OM 600 2 color R. Kennedy and C. Taylor. Optical flow with geometric occlusion estimation and fusion of multiple frames. EMMCVPR 2015.
[101] PH-Flow 800 2 color J. Yang and H. Li. Dense, accurate optical flow estimation with piecewise parametric model. CVPR 2015.
[102] EpicFlow 16 2 color J. Revaud, P. Weinzaepfel, Z. Harchaoui, and C. Schmid. EpicFlow: edge-preserving interpolation of correspondences for optical flow. CVPR 2015.
[103] NNF-EAC 380 2 color Anonymous. Variational method for joint optical flow estimation and edge-aware image restoration. CVPR 2015 submission 2336.
[104] Heeger++ 6600 5 gray Anonymous. A context aware biologically inspired algorithm for optical flow (updated results). CVPR 2015 submission 2238.
[105] HBM-GC 330 2 color A. Zheng and Y. Yuan. Motion estimation via hierarchical block matching and graph cut. Submitted to ICIP 2015.
[106] FFV1MT 358 5 gray F. Solari, M. Chessa, N. Medathati, and P. Kornprobst. What can we expect from a V1-MT feedforward architecture for optical flow estimation? Submitted to Signal Processing: Image Communication 2015.
[107] ROF-ND 4 2 color S. Ali, C. Daul, E. Galbrun, and W. Blondel. Illumination invariant large displacement optical flow using robust neighbourhood descriptors. Submitted to CVIU 2015.
[108] DeepFlow2 16 2 color J. Revaud, P. Weinzaepfel, Z. Harchaoui, and C. Schmid. Deep convolutional matching. Submitted to IJCV, 2015.
[109] HAST 2667 2 color Anonymous. Highly accurate optical flow estimation on superpixel tree. ICCV 2015 submission 2221.
[110] FlowFields 15 2 color Anonymous. Flow Fields: Dense unregularized correspondence fields for highly accurate large displacement optical flow estimation. ICCV 2015 submission 744.
[111] SVFilterOh 1.56 2 color Anonymous. Fast estimation of large displacement optical flow using PatchMatch and dominant motion patterns. CVPR 2016 submission 1788.
[112] FlowNetS+ft+v 0.5 2 color Anonymous. Learning optical flow with convolutional neural networks. ICCV 2015 submission 235.
[113] CombBMOF 51 2 color M. Brüggemann, R. Kays, P. Springer, and O. Erdler. Combined block-matching and adaptive differential motion estimation in a hierarchical multi-scale framework. ICGIP 2014. (Method improved since publication.)
[114] PMMST 182 2 color F. Zhang, S. Xu, and X. Zhang. High accuracy correspondence field estimation via MST based patch matching. Submitted to TIP 2015.
[115] DF-Auto 70 2 color N. Monzon, A. Salgado, and J. Sanchez. Regularization strategies for discontinuity-preserving optical flow methods. Submitted to TIP 2015.
[116] CPM-Flow 3 2 color Anonymous. Efficient coarse-to-fine PatchMatch for large displacement optical flow. CVPR 2016 submission 241.
[117] CNN-flow-warp+ref 1.4 3 color D. Teney and M. Hebert. Learning to extract motion from videos in convolutional neural networks. ArXiv 1601.07532, 2016.
[118] Steered-L1 804 2 color Anonymous. Optical flow estimation via steered-L1 norm. Submitted to WSCG 2016.
[119] StereoOF-V1MT 343 2 gray Anonymous. Visual features for action-oriented tasks: a cortical-like model for disparity and optic flow computation. BMVC 2016 submission 132.
[120] PGM-C 5 2 color Y. Li. Pyramidal gradient matching for optical flow estimation. Submitted to PAMI 2016.
[121] RNLOD-Flow 1040 2 gray C. Zhang, Z. Chen, M. Wang, M. Li, and S. Jiang. Robust non-local TV-L1 optical flow estimation with occlusion detection. Submitted to TIP 2016.
[122] FlowNet2 0.091 2 color Anonymous. FlowNet 2.0: Evolution of optical flow estimation with deep networks. CVPR 2017 submission 900.
[123] S2F-IF 20 2 color Anonymous. S2F-IF: Slow-to-fast interpolator flow. CVPR 2017 submission 765.
[124] BriefMatch 0.068 2 gray G. Eilertsen, P.-E. Forssen, and J. Unger. Dense binary feature matching for real-time optical flow estimation. SCIA 2017 submission 62.
[125] OAR-Flow 60 2 color Anonymous. Order-adaptive regularisation for variational optical flow: global, local and in between. SSVM 2017 submission 20.
[126] AdaConv-v1 2.8 2 color S. Niklaus, L. Mai, and F. Liu. (Interpolation results only.) Video frame interpolation via adaptive convolution. CVPR 2017.
[127] SepConv-v1 0.2 2 color S. Niklaus, L. Mai, and F. Liu. (Interpolation results only.) Video frame interpolation via adaptive separable convolution. ICCV 2017.
[128] ProbFlowFields 37 2 color A. Wannenwetsch, M. Keuper, and S. Roth. ProbFlow: joint optical flow and uncertainty estimation. ICCV 2017.
[129] UnFlow 0.12 2 color Anonymous. UnFlow: Unsupervised learning of optical flow with a bidirectional census loss. Submitted to AAAI 2018.
* The "time" column lists the reported runtime in seconds on the "Urban" sequence. Note that these runtimes are not normalized by processor speed or type.