Optical flow evaluation results Statistics:     Average   SD   R0.5   R1.0   R2.0   A50   A75   A95  
Error type:   endpoint   angle   interpolation   normalized interpolation  
Show images: below table   above table        
R0.5
endpoint
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
OFLAF [77]10.5 1.67 4 9.54 5 0.81 11 4.12 7 23.1 5 2.26 15 3.43 1 11.8 1 1.47 7 2.31 8 15.3 5 0.67 8 18.1 2 27.2 3 8.68 5 12.6 34 27.0 11 6.93 28 0.78 16 1.08 27 3.37 22 11.6 8 25.9 3 16.2 15
MDP-Flow2 [68]12.2 1.76 8 9.93 9 0.86 13 3.26 2 20.6 2 1.44 3 4.04 3 13.6 3 1.26 5 3.09 26 21.8 30 0.88 19 22.4 15 32.2 12 14.3 14 9.15 11 23.6 3 5.83 10 2.18 38 0.32 15 4.78 35 10.8 5 26.3 4 15.0 9
NNF-Local [87]12.9 1.75 7 9.66 6 0.87 14 4.56 14 27.6 16 2.46 19 4.52 4 15.6 7 2.18 16 2.57 15 18.4 15 0.86 16 18.8 5 29.2 5 8.22 3 9.37 14 24.2 6 5.22 7 0.95 18 2.02 60 1.90 13 10.6 4 31.8 24 8.21 2
NN-field [71]15.1 2.01 19 11.0 20 1.10 27 5.74 26 30.7 31 3.21 27 4.57 5 15.7 8 2.22 18 1.81 2 15.7 7 0.54 5 19.0 6 29.4 6 8.42 4 10.5 20 20.3 1 3.83 1 1.43 24 2.03 61 3.72 23 10.1 2 31.1 19 6.81 1
PMMST [114]17.0 1.43 2 8.08 2 0.38 1 6.05 32 28.6 20 4.40 40 6.13 17 19.1 16 4.02 39 2.21 5 11.7 1 1.09 27 21.2 9 30.5 9 13.2 11 9.59 16 23.8 4 5.71 9 2.28 41 1.95 55 4.36 29 11.8 9 29.1 8 13.1 6
WLIF-Flow [93]20.6 1.80 10 9.97 10 0.94 16 6.24 35 28.7 22 4.52 42 5.74 12 18.5 11 3.13 25 3.03 25 19.9 22 1.01 24 21.9 10 32.0 11 14.1 12 12.4 31 27.2 12 6.44 13 3.40 62 0.07 6 8.69 60 11.4 7 26.8 5 15.7 12
ComponentFusion [96]20.8 1.73 6 9.92 8 0.77 8 3.75 5 22.3 4 2.27 16 5.02 9 17.1 10 2.21 17 2.87 19 19.3 20 0.93 20 24.1 19 35.1 22 18.6 34 12.4 31 39.0 70 8.29 54 2.41 42 0.13 13 4.73 34 11.9 11 30.0 15 15.5 11
Correlation Flow [75]21.7 1.96 17 10.9 18 0.66 5 4.21 9 25.5 12 1.35 2 6.69 29 20.7 24 0.94 2 1.68 1 13.3 2 0.48 4 25.4 27 36.9 29 16.3 24 13.5 50 32.5 35 7.88 48 2.82 51 1.63 42 10.8 68 11.3 6 29.7 11 9.89 4
NNF-EAC [103]23.2 1.95 16 10.5 15 1.10 27 4.29 11 25.1 10 2.20 12 5.06 10 16.6 9 1.81 11 3.94 52 23.4 39 1.63 52 22.5 16 32.5 15 14.7 16 11.3 24 25.8 8 6.81 22 2.71 48 2.13 63 4.63 33 12.4 14 30.2 17 16.5 16
Layers++ [37]24.1 1.85 13 10.1 11 1.03 23 6.26 36 27.9 19 4.58 44 4.88 8 15.2 5 3.65 33 2.26 7 14.4 3 0.68 9 17.8 1 25.4 1 12.2 8 13.3 44 28.3 16 6.81 22 4.53 71 2.71 80 7.34 51 13.2 22 29.7 11 19.2 40
LME [70]24.2 1.89 15 10.6 16 0.75 6 3.53 3 21.3 3 1.83 8 7.04 40 18.6 14 8.24 71 3.10 28 23.0 37 0.86 16 24.8 25 35.5 24 17.9 31 10.1 18 30.1 26 6.46 14 2.67 47 1.51 39 5.54 39 12.8 16 30.9 18 17.7 27
RNLOD-Flow [121]26.2 1.68 5 9.49 4 0.75 6 5.76 27 30.6 30 3.18 25 6.58 26 21.1 28 2.92 22 2.61 16 17.5 12 0.74 11 22.0 11 32.8 16 14.9 19 11.5 26 28.0 15 7.37 36 5.53 85 4.17 93 15.2 91 11.8 9 27.7 7 15.1 10
nLayers [57]26.4 1.40 1 7.64 1 0.64 3 8.47 67 30.5 29 7.07 78 6.80 35 20.0 20 5.79 63 2.13 4 14.7 4 0.78 14 18.3 3 26.0 2 12.2 8 12.9 39 25.6 7 6.84 24 1.93 34 2.24 67 3.35 21 14.5 34 32.0 27 20.9 49
HAST [109]27.7 1.57 3 8.65 3 0.60 2 5.82 30 24.9 9 3.84 32 3.85 2 12.5 2 0.41 1 2.82 18 18.8 17 0.58 6 18.7 4 27.9 4 8.00 1 15.9 85 32.6 36 9.69 76 9.89 110 4.87 102 36.9 117 8.55 1 22.0 1 9.33 3
FC-2Layers-FF [74]28.4 1.81 11 9.71 7 1.07 26 6.99 44 33.9 45 4.71 45 4.71 6 15.0 4 3.63 32 2.67 17 18.3 14 0.87 18 20.6 8 29.7 7 14.5 15 13.3 44 28.4 17 7.26 33 5.67 89 1.82 50 14.9 87 12.9 19 29.4 9 18.4 34
SVFilterOh [111]29.3 2.18 29 11.4 24 0.78 9 5.94 31 29.2 24 3.13 23 4.85 7 15.2 5 2.01 13 2.35 9 17.4 10 0.63 7 20.4 7 30.1 8 8.12 2 14.7 77 29.5 22 8.51 61 10.3 111 3.98 91 31.3 109 12.0 12 26.9 6 13.4 7
TC/T-Flow [76]29.6 2.19 31 11.8 31 1.21 36 5.00 17 29.4 25 1.89 9 5.57 11 18.5 11 1.28 6 3.63 41 23.7 42 1.24 34 24.1 19 35.4 23 15.2 21 7.02 2 25.8 8 5.43 8 3.79 66 2.13 63 19.3 100 14.6 37 36.2 40 17.9 29
AGIF+OF [85]30.1 1.99 18 10.9 18 1.16 32 8.96 72 37.8 65 6.95 73 6.31 20 20.4 21 3.86 37 2.89 20 19.1 18 0.93 20 22.1 12 32.4 13 14.8 17 12.9 39 28.6 19 6.79 21 3.48 65 0.08 11 8.67 59 12.8 16 29.9 14 17.5 23
TC-Flow [46]30.5 2.04 20 11.0 20 0.94 16 3.69 4 23.4 6 1.68 5 5.96 16 19.9 19 1.06 3 3.73 44 23.6 41 1.24 34 25.7 31 38.0 34 14.8 17 8.99 10 34.6 45 5.03 6 2.82 51 2.47 73 15.9 95 16.0 44 38.1 46 22.0 53
HBM-GC [105]31.2 2.10 24 10.8 17 0.94 16 7.69 57 31.9 38 6.10 66 6.21 19 18.5 11 3.82 35 2.22 6 15.5 6 0.78 14 22.1 12 31.8 10 14.2 13 13.3 44 24.1 5 7.49 40 8.91 107 1.87 51 21.4 104 12.7 15 31.1 19 16.7 19
ALD-Flow [66]32.8 2.22 33 11.8 31 1.02 22 4.33 12 24.7 7 2.04 11 6.35 22 20.9 25 1.56 8 3.67 42 24.0 46 1.10 28 25.8 32 37.5 33 15.8 23 8.24 5 32.7 38 4.68 3 3.06 56 2.62 79 16.6 97 15.6 43 39.3 48 19.8 44
IROF++ [58]35.2 2.18 29 11.5 27 1.33 43 7.91 58 36.5 60 5.96 63 6.71 31 21.7 34 4.85 56 3.62 40 22.8 33 1.63 52 24.2 22 34.9 21 17.1 28 13.3 44 33.4 41 7.82 46 0.55 13 1.09 28 1.01 9 12.8 16 32.9 31 16.7 19
ProbFlowFields [128]35.2 3.29 69 17.1 72 1.88 83 6.54 38 29.8 26 5.50 56 7.69 47 23.8 45 6.90 66 3.09 26 17.1 9 1.28 36 27.8 39 39.5 40 19.2 40 6.56 1 27.2 12 5.00 5 0.09 3 0.03 2 0.86 7 17.1 51 39.4 49 17.5 23
FESL [72]35.7 1.86 14 10.2 12 0.99 19 10.4 86 39.3 76 7.94 85 6.79 34 21.2 30 4.35 46 2.39 10 16.2 8 0.76 13 24.2 22 34.7 19 19.2 40 12.6 34 27.7 14 7.10 32 3.35 60 2.20 66 8.03 55 14.5 34 31.2 21 17.6 26
FMOF [94]36.1 2.09 22 11.2 23 1.44 56 9.20 74 37.9 66 6.96 74 6.14 18 19.5 17 3.82 35 2.52 13 17.4 10 0.73 10 24.1 19 34.8 20 17.7 30 13.5 50 28.4 17 6.99 29 4.64 74 1.63 42 14.5 83 14.5 34 32.7 29 17.2 22
Classic+CPF [83]36.8 2.16 27 11.6 28 1.42 54 8.04 60 36.6 61 5.76 60 6.70 30 21.8 36 3.86 37 3.02 24 21.2 28 1.14 30 23.6 18 34.2 18 17.0 27 13.4 49 29.1 20 7.03 31 4.58 73 1.50 38 13.6 80 12.9 19 29.7 11 17.5 23
MLDP_OF [89]37.0 2.46 45 13.6 48 1.14 31 4.43 13 27.8 18 1.99 10 6.80 35 21.7 34 1.89 12 2.52 13 19.6 21 0.75 12 28.2 40 40.4 43 19.2 40 11.8 28 29.6 23 9.40 72 8.30 105 2.15 65 31.7 111 13.3 23 33.4 34 15.9 13
PH-Flow [101]37.6 2.30 38 12.2 35 1.45 58 7.61 55 35.2 50 5.81 61 5.92 15 19.0 15 4.67 54 3.52 36 22.1 31 1.49 46 23.4 17 33.9 17 16.3 24 12.4 31 29.4 21 6.87 25 4.88 78 2.25 69 14.0 81 12.2 13 30.1 16 16.6 17
Efficient-NL [60]38.9 2.41 43 11.8 31 1.55 69 8.28 63 35.4 53 5.90 62 6.72 32 20.9 25 3.78 34 2.95 23 19.1 18 1.20 32 22.1 12 32.4 13 15.3 22 14.6 75 31.0 30 8.05 52 3.32 58 2.40 72 7.02 47 13.6 27 29.6 10 18.2 31
PMF [73]39.3 2.30 38 12.8 41 1.01 20 5.64 25 31.4 35 2.51 20 6.83 37 22.5 41 1.72 10 3.26 29 21.1 26 0.94 22 24.3 24 36.7 28 9.56 6 14.4 71 40.0 76 8.40 58 7.45 102 8.62 119 24.8 105 10.4 3 25.6 2 12.7 5
Aniso-Texture [82]39.9 1.84 12 10.4 14 0.81 11 4.68 15 24.7 7 3.74 30 8.07 51 23.8 45 3.23 28 2.10 3 18.4 15 0.46 2 29.0 51 41.0 48 22.9 55 12.7 36 32.7 38 8.37 57 6.63 97 5.87 110 14.8 86 16.4 45 35.2 38 24.2 63
OAR-Flow [125]40.3 2.96 63 15.1 63 1.58 71 6.72 41 31.3 34 4.10 37 9.12 59 27.1 56 4.72 55 3.73 44 23.9 44 1.17 31 28.3 42 40.8 45 17.6 29 8.30 6 33.5 43 4.69 4 0.25 7 0.17 14 2.56 17 17.5 55 40.5 51 22.6 56
Sparse-NonSparse [56]41.0 2.11 25 11.7 29 1.39 49 7.47 54 35.1 49 5.75 59 6.48 24 21.2 30 4.34 44 3.52 36 22.9 34 1.38 41 26.1 35 37.3 32 19.6 47 13.5 50 31.2 32 7.42 37 5.02 82 1.18 32 13.5 79 13.5 25 31.7 22 18.8 38
Ramp [62]41.8 2.21 32 12.1 34 1.45 58 7.45 53 35.5 55 5.63 58 6.33 21 20.6 23 4.23 41 3.43 33 22.5 32 1.38 41 25.8 32 37.0 30 19.3 44 13.5 50 30.3 27 7.45 38 4.89 79 1.97 57 15.1 90 13.1 21 31.9 25 18.0 30
NL-TV-NCC [25]41.9 2.25 35 11.7 29 0.78 9 6.94 43 35.5 55 2.54 21 6.48 24 21.1 28 1.08 4 2.48 12 21.5 29 0.46 2 31.5 64 46.7 81 16.7 26 17.3 89 41.0 85 10.2 79 4.41 70 0.10 12 10.1 65 18.4 56 43.2 57 18.2 31
LSM [39]42.2 2.24 34 12.4 37 1.39 49 7.37 51 35.4 53 5.47 55 6.61 27 21.6 33 4.24 42 3.46 35 23.8 43 1.30 38 25.8 32 37.0 30 19.3 44 13.7 56 32.1 34 7.36 35 5.34 84 1.11 30 14.5 83 13.7 30 32.3 28 18.2 31
OFH [38]42.6 2.82 58 13.9 51 2.01 84 4.91 16 28.6 20 2.33 17 8.97 58 28.0 61 2.88 21 4.00 55 27.0 56 1.43 44 31.0 61 44.5 68 22.7 53 10.5 20 41.8 87 6.88 26 0.03 1 0.02 1 0.27 3 17.1 51 46.4 70 19.2 40
Sparse Occlusion [54]43.0 2.14 26 11.4 24 1.03 23 7.32 49 31.0 32 6.11 67 7.29 44 22.9 43 2.48 20 3.29 30 22.9 34 1.03 25 26.9 37 39.4 38 14.9 19 13.0 41 33.3 40 7.63 43 7.80 103 8.76 122 12.2 75 14.8 39 34.8 37 16.9 21
Classic+NL [31]43.2 2.08 21 11.4 24 1.35 45 7.33 50 35.9 58 5.30 52 6.47 23 21.0 27 4.53 51 3.59 38 22.9 34 1.49 46 25.4 27 36.3 26 19.4 46 13.8 59 31.1 31 7.57 41 5.78 91 2.32 71 15.0 88 13.5 25 31.9 25 18.7 37
IROF-TV [53]45.7 2.51 48 13.5 46 1.41 52 8.08 61 38.7 73 6.19 69 6.97 39 22.3 38 4.43 48 4.23 58 28.8 64 1.72 57 28.3 42 39.9 41 22.6 52 13.8 59 40.0 76 8.01 50 0.23 6 0.39 19 0.67 5 13.7 30 33.6 35 17.8 28
RFlow [90]45.8 2.42 44 13.5 46 1.16 32 3.98 6 25.2 11 1.81 7 8.89 57 27.5 58 3.13 25 3.45 34 26.8 54 1.60 50 30.5 59 43.6 59 24.5 63 14.2 66 38.1 65 7.94 49 3.36 61 1.65 45 8.47 58 16.9 50 42.3 53 20.5 48
TV-L1-MCT [64]45.9 2.09 22 11.1 22 1.39 49 9.67 76 39.0 75 7.35 79 7.11 42 22.3 38 4.27 43 2.94 22 20.7 25 1.13 29 28.4 44 39.4 38 25.8 70 16.0 87 35.0 48 9.27 69 1.27 22 0.57 21 7.24 49 14.8 39 33.2 33 23.3 60
COFM [59]46.5 2.51 48 13.9 51 1.42 54 5.54 21 28.9 23 3.40 28 7.79 49 23.6 44 4.53 51 2.90 21 18.2 13 0.95 23 29.1 52 40.8 45 25.4 67 15.5 83 30.7 29 9.39 71 4.69 75 1.23 33 15.4 94 16.8 49 37.1 41 21.8 52
S2D-Matching [84]46.7 2.39 41 13.0 42 1.52 66 7.22 47 35.6 57 5.13 50 7.63 46 24.4 48 4.38 47 3.30 31 20.5 24 1.35 40 25.3 26 36.0 25 19.2 40 14.1 63 31.5 33 7.77 45 6.21 94 2.24 67 16.9 98 13.6 27 31.7 22 19.3 42
Complementary OF [21]48.4 2.69 55 15.0 61 1.33 43 4.19 8 26.9 14 1.70 6 7.15 43 24.4 48 3.09 24 3.86 50 26.1 50 1.41 43 33.2 76 44.6 70 29.7 85 13.3 44 40.7 84 7.00 30 0.74 15 0.03 2 7.12 48 23.9 83 53.1 90 33.9 89
ACK-Prior [27]48.5 2.16 27 12.4 37 0.64 3 4.26 10 25.9 13 1.33 1 5.91 14 20.4 21 1.67 9 2.39 10 20.4 23 0.33 1 30.0 56 41.1 49 24.4 62 19.2 100 40.3 81 12.2 92 14.4 120 6.30 113 40.9 120 21.0 68 43.6 59 27.5 75
Occlusion-TV-L1 [63]48.6 2.47 46 13.1 43 1.12 29 6.32 37 32.2 40 4.45 41 9.86 64 28.3 62 4.44 49 3.86 50 26.7 53 1.43 44 31.9 70 44.3 65 27.1 76 11.3 24 35.3 52 10.4 81 0.47 12 1.16 31 0.67 5 19.8 63 46.6 72 22.9 58
CostFilter [40]48.8 2.65 52 14.9 59 1.01 20 5.51 20 31.6 36 2.23 14 7.39 45 24.1 47 3.06 23 3.82 48 26.1 50 1.08 26 25.4 27 38.9 37 10.2 7 15.1 80 42.7 90 8.52 62 8.99 108 10.3 124 29.5 108 14.4 33 37.2 42 16.1 14
MDP-Flow [26]49.0 2.33 40 13.4 45 1.20 34 5.61 24 26.9 14 4.88 47 6.76 33 22.6 42 5.72 62 4.13 56 29.5 69 1.92 61 28.7 47 40.8 45 23.2 58 12.8 38 36.7 57 8.04 51 2.54 44 2.96 82 4.43 31 19.4 62 44.6 63 26.0 70
2DHMM-SAS [92]50.7 2.28 37 12.3 36 1.46 60 8.45 66 38.1 68 6.02 65 8.34 53 24.4 48 5.21 58 3.73 44 23.4 39 1.62 51 25.5 30 36.6 27 18.8 35 14.5 72 33.4 41 8.46 60 5.07 83 2.05 62 15.3 92 13.3 23 32.9 31 18.5 36
SimpleFlow [49]51.2 2.39 41 12.6 40 1.60 73 9.03 73 38.3 69 7.38 80 8.52 55 25.7 52 5.41 59 4.36 63 25.5 49 2.48 69 26.8 36 38.0 34 21.6 49 14.0 62 29.7 24 7.65 44 2.77 50 1.97 57 6.48 44 14.3 32 32.8 30 19.7 43
Steered-L1 [118]52.1 1.78 9 10.2 12 0.92 15 2.76 1 19.0 1 1.44 3 5.78 13 19.7 18 2.10 14 3.99 54 27.5 59 1.53 49 31.3 63 43.3 57 27.6 78 14.6 75 39.2 72 9.88 77 14.6 121 5.70 108 47.1 121 22.4 75 48.1 75 29.3 80
AggregFlow [97]53.5 3.49 77 17.6 76 1.74 75 10.4 86 42.0 87 6.99 75 11.4 74 30.6 67 9.73 79 3.75 47 21.1 26 1.52 48 28.9 50 41.8 53 18.2 32 7.46 3 22.8 2 4.30 2 1.95 35 2.54 75 4.24 27 20.5 65 42.7 56 25.8 68
TF+OM [100]54.0 2.89 62 14.7 57 1.35 45 5.44 19 27.6 16 3.94 33 10.2 65 26.3 53 12.8 88 3.61 39 24.6 48 1.28 36 29.8 55 40.5 44 25.7 68 12.7 36 34.9 47 6.11 11 5.62 88 4.89 103 14.0 81 21.5 70 46.5 71 23.8 62
EPPM w/o HM [88]56.0 3.53 79 16.6 68 1.44 56 5.80 28 35.3 52 2.22 13 8.01 50 26.3 53 2.45 19 4.38 64 28.0 60 1.77 59 27.0 38 40.0 42 13.0 10 18.6 95 44.4 98 10.8 84 10.5 113 2.30 70 37.7 118 13.6 27 35.6 39 13.9 8
DeepFlow2 [108]56.1 3.18 67 16.9 71 1.48 62 6.68 40 33.7 44 4.15 38 10.7 69 31.1 68 7.69 69 5.96 80 31.1 75 3.33 79 28.5 45 41.3 50 19.1 37 9.45 15 34.7 46 6.60 16 1.60 27 1.66 47 10.4 66 23.5 79 47.9 74 30.2 83
CombBMOF [113]56.5 2.67 53 14.4 54 1.05 25 7.17 45 33.9 45 4.00 34 6.89 38 21.4 32 4.18 40 5.20 73 28.8 64 3.04 77 28.2 40 41.7 52 18.5 33 21.8 104 39.5 73 20.1 108 4.02 67 4.61 100 6.08 41 17.1 51 37.5 43 24.5 65
Adaptive [20]57.0 2.49 47 13.2 44 1.13 30 7.17 45 34.5 48 4.97 48 10.2 65 28.7 64 4.34 44 4.31 62 28.2 61 1.66 54 34.5 83 48.1 87 28.2 80 14.3 69 36.1 54 8.24 53 4.04 68 4.49 99 7.32 50 14.7 38 34.7 36 18.9 39
ComplOF-FED-GPU [35]59.4 2.87 61 15.6 65 1.35 45 6.14 33 33.6 43 3.15 24 8.26 52 27.4 57 3.30 29 4.29 59 28.2 61 1.71 56 33.1 75 47.9 86 24.1 61 14.7 77 45.7 101 9.19 68 3.33 59 1.48 37 15.0 88 18.8 59 48.4 77 22.0 53
TCOF [69]59.7 3.05 65 15.4 64 1.75 76 8.12 62 38.6 71 5.20 51 13.8 84 34.5 81 13.2 90 8.75 96 29.0 66 8.90 101 33.8 79 47.3 83 23.1 57 9.33 13 25.9 10 6.64 18 2.59 45 1.95 55 6.38 43 15.3 42 39.1 47 18.4 34
ROF-ND [107]60.7 3.29 69 14.7 57 1.29 40 7.42 52 31.8 37 2.42 18 7.10 41 22.1 37 2.13 15 3.68 43 23.2 38 2.87 75 31.1 62 43.7 60 23.6 60 19.3 102 38.5 67 10.7 83 8.90 106 2.99 83 24.8 105 21.9 72 48.6 78 22.7 57
BriefMatch [124]60.8 2.27 36 12.5 39 1.23 38 6.23 34 32.1 39 3.54 29 6.68 28 22.3 38 3.16 27 3.32 32 23.9 44 1.23 33 30.8 60 43.3 57 26.8 73 23.9 107 43.6 93 21.0 110 11.0 116 4.44 98 33.1 114 21.5 70 44.8 64 29.2 79
DeepFlow [86]61.1 3.36 72 17.3 73 1.54 68 7.91 58 35.2 50 5.35 53 12.1 78 33.0 76 10.5 84 6.24 82 32.1 77 3.55 81 28.8 48 42.3 54 18.8 35 9.77 17 37.4 61 6.90 27 1.30 23 0.37 16 9.70 64 27.4 93 52.7 86 35.8 91
Classic++ [32]61.6 2.59 51 13.9 51 1.51 65 6.79 42 32.3 41 5.37 54 9.15 60 27.8 60 5.54 60 4.29 59 29.0 66 1.77 59 30.2 58 43.9 61 22.8 54 14.8 79 40.0 76 8.36 56 6.82 98 4.17 93 16.5 96 16.5 46 39.5 50 19.8 44
TV-L1-improved [17]62.6 2.56 50 13.6 48 1.20 34 5.80 28 30.0 28 4.04 35 9.84 63 28.4 63 4.60 53 4.16 57 27.0 56 1.66 54 31.5 64 45.4 75 23.0 56 17.5 90 45.5 100 13.7 97 7.01 100 4.32 97 20.5 102 17.1 51 42.5 55 20.4 47
Bartels [41]63.8 3.26 68 16.7 70 1.37 48 5.33 18 29.8 26 3.18 25 8.40 54 26.9 55 4.45 50 4.40 65 26.9 55 2.16 63 32.7 72 45.4 75 28.4 82 14.3 69 38.1 65 12.9 93 8.20 104 3.82 89 31.3 109 18.4 56 43.7 61 23.3 60
S2F-IF [123]64.3 4.50 93 23.8 98 2.14 89 8.86 69 42.8 91 6.52 70 11.0 70 34.0 80 9.17 74 4.86 68 29.2 68 2.39 67 35.6 89 51.0 100 26.9 74 8.49 8 36.6 56 6.30 12 0.60 14 0.03 2 2.49 16 23.6 80 52.7 86 25.9 69
SIOF [67]64.4 2.67 53 13.6 48 1.23 38 7.65 56 37.9 66 4.78 46 14.2 87 32.9 75 15.4 92 6.36 83 34.7 83 3.84 83 34.0 80 45.8 78 33.2 89 13.5 50 35.7 53 10.5 82 1.99 36 0.99 25 4.21 26 20.5 65 45.6 66 30.7 85
F-TV-L1 [15]65.6 3.34 71 17.3 73 1.80 80 9.99 83 38.6 71 7.03 76 13.2 82 32.6 74 7.82 70 5.85 79 32.9 78 2.91 76 31.5 64 45.0 72 25.2 65 15.1 80 38.7 68 8.99 66 2.19 39 3.29 86 3.03 19 15.1 41 38.0 45 16.6 17
CRTflow [80]66.2 3.53 79 18.6 80 1.79 78 6.54 38 34.0 47 4.08 36 10.5 67 31.6 71 5.02 57 4.95 71 30.6 72 2.31 64 30.1 57 44.2 64 19.1 37 24.2 108 50.1 107 26.0 114 1.80 32 0.92 24 6.63 45 22.7 76 52.1 84 30.0 82
PGM-C [120]66.9 4.80 97 24.9 103 2.34 96 9.79 77 42.4 88 7.86 83 11.3 73 34.5 81 9.49 75 5.28 76 34.6 82 2.54 72 34.5 83 49.1 93 26.5 71 9.26 12 37.2 58 6.72 20 0.42 10 0.07 6 1.85 12 23.7 81 53.0 89 25.6 67
FlowFields [110]68.0 4.67 94 24.4 100 2.22 91 9.80 78 44.3 95 7.67 81 11.8 76 36.4 88 10.1 82 4.90 70 30.5 71 2.63 73 36.4 91 51.6 103 28.9 84 8.76 9 38.7 68 6.48 15 0.85 17 0.03 2 2.71 18 23.3 78 53.7 93 22.3 55
TriangleFlow [30]68.2 2.81 57 14.9 59 1.22 37 7.27 48 37.1 63 3.76 31 9.83 62 30.2 66 3.34 30 3.84 49 27.1 58 1.72 57 39.4 104 53.7 106 34.8 94 21.8 104 43.5 92 16.0 100 4.72 77 7.40 116 8.30 57 18.5 58 44.3 62 21.5 51
CPM-Flow [116]68.9 4.79 96 24.9 103 2.32 94 9.83 79 42.4 88 7.89 84 11.2 72 33.9 79 9.50 76 5.25 74 34.3 80 2.50 71 34.7 86 49.3 96 26.7 72 10.3 19 37.5 63 7.62 42 0.42 10 0.07 6 1.82 11 24.5 85 54.2 94 26.8 73
Rannacher [23]69.1 3.03 64 16.1 66 1.59 72 8.35 64 36.9 62 6.87 72 11.1 71 31.8 72 6.71 65 4.88 69 29.7 70 2.34 65 31.7 69 45.9 79 23.3 59 16.8 88 44.0 95 10.3 80 4.89 79 2.57 77 12.1 74 16.7 48 41.9 52 19.9 46
LocallyOriented [52]69.1 4.06 87 20.2 88 1.87 82 12.1 90 47.6 98 8.49 88 15.9 91 39.1 96 11.1 86 5.10 72 28.6 63 2.84 74 34.0 80 47.4 84 25.7 68 11.8 28 32.6 36 7.84 47 1.10 19 1.51 39 6.95 46 20.3 64 46.8 73 23.1 59
SRR-TVOF-NL [91]69.1 3.16 66 16.2 67 1.49 63 8.87 70 38.5 70 5.57 57 12.3 79 33.4 77 8.53 72 3.96 53 26.6 52 1.34 39 32.8 73 44.6 70 27.2 77 13.8 59 39.0 70 8.34 55 5.55 87 5.38 107 17.8 99 22.0 73 43.3 58 25.2 66
EpicFlow [102]69.4 4.80 97 24.9 103 2.33 95 9.90 81 42.9 92 7.95 86 11.8 76 35.7 86 9.56 77 5.26 75 34.4 81 2.49 70 34.6 85 49.2 95 27.0 75 10.8 23 37.5 63 7.33 34 0.41 9 0.07 6 1.80 10 24.1 84 53.5 92 26.4 71
Aniso. Huber-L1 [22]69.6 2.84 60 14.5 55 1.46 60 14.0 92 42.6 90 12.9 91 13.4 83 31.3 69 13.0 89 6.50 84 35.2 85 4.19 86 29.7 54 42.4 55 21.8 51 14.5 72 35.0 48 8.43 59 5.54 86 3.18 85 12.8 77 16.6 47 37.7 44 20.9 49
Dynamic MRF [7]70.9 3.39 73 18.9 83 1.30 41 5.60 23 33.5 42 2.81 22 9.67 61 31.3 69 3.54 31 4.64 67 33.7 79 2.39 67 38.0 97 51.2 101 34.9 95 19.2 100 51.8 110 15.2 99 3.41 63 0.37 16 20.9 103 25.1 87 52.2 85 31.7 87
DPOF [18]73.4 4.03 85 21.8 92 2.11 86 9.50 75 40.4 79 5.97 64 8.88 56 27.5 58 6.05 64 4.29 59 30.8 73 2.08 62 31.5 64 45.1 73 21.6 49 15.9 85 37.3 60 9.53 74 15.3 122 1.61 41 47.3 122 22.1 74 46.3 69 28.5 76
CBF [12]75.5 2.82 58 15.0 61 1.32 42 18.0 97 40.4 79 21.6 99 10.6 68 29.2 65 9.72 78 6.57 85 34.8 84 4.55 88 31.5 64 44.0 63 24.5 63 14.5 72 35.0 48 8.92 65 10.9 115 6.02 111 26.2 107 20.7 67 43.6 59 27.2 74
Brox et al. [5]75.6 3.55 82 18.8 82 1.64 74 10.1 85 39.6 77 8.97 89 11.7 75 33.4 77 8.96 73 6.57 85 36.4 87 3.41 80 38.2 99 47.6 85 45.3 111 13.5 50 42.4 89 9.63 75 0.27 8 0.99 25 0.47 4 31.0 99 56.4 100 43.3 103
Fusion [6]75.6 3.40 74 19.1 84 2.16 90 5.57 22 31.1 33 4.53 43 7.70 48 25.2 51 7.53 68 5.78 78 35.6 86 4.10 85 36.6 92 47.1 82 38.8 100 14.2 66 41.9 88 13.2 94 6.84 99 5.31 105 11.7 72 24.8 86 51.1 83 31.6 86
Local-TV-L1 [65]76.4 4.05 86 19.4 86 2.51 98 17.1 96 43.6 94 15.9 94 19.8 98 37.3 91 23.3 95 9.20 99 43.3 95 6.89 98 28.6 46 41.3 50 20.2 48 14.1 63 35.1 51 8.67 63 1.24 20 0.62 22 3.94 24 33.5 106 57.2 101 49.6 110
CLG-TV [48]77.2 2.80 56 14.6 56 1.41 52 14.0 92 40.7 82 14.1 93 12.7 80 32.0 73 11.0 85 8.13 92 47.7 103 5.99 94 32.0 71 45.2 74 25.2 65 14.1 63 40.1 80 10.9 85 6.45 95 5.82 109 10.4 66 19.0 60 42.4 54 26.6 72
LDOF [28]78.8 4.09 89 20.0 87 2.31 93 9.96 82 41.8 85 7.06 77 14.1 85 37.0 89 10.1 82 8.41 93 43.3 95 4.97 89 34.4 82 46.2 80 32.5 87 12.2 30 41.0 85 8.88 64 1.63 29 2.00 59 5.79 40 29.9 94 56.0 98 38.8 98
p-harmonic [29]78.9 3.47 76 19.1 84 2.29 92 8.40 65 35.9 58 6.80 71 12.8 81 34.6 83 9.84 81 9.04 97 47.6 102 6.72 96 37.1 94 48.7 91 39.6 101 13.1 42 44.0 95 11.2 86 3.43 64 2.50 74 6.33 42 21.2 69 45.2 65 30.5 84
DF-Auto [115]79.0 4.71 95 22.2 93 2.11 86 21.1 100 49.4 100 21.6 99 20.3 99 39.8 98 31.0 101 7.62 89 37.5 89 5.08 91 33.6 77 43.9 61 33.3 90 8.36 7 29.8 25 7.48 39 2.60 46 5.21 104 2.22 14 32.4 102 53.1 90 43.2 102
TriFlow [95]80.8 3.53 79 17.9 77 1.77 77 11.0 89 37.3 64 10.6 90 16.7 94 35.8 87 25.3 97 4.44 66 30.9 74 2.34 65 35.0 88 44.3 65 35.7 96 10.7 22 30.4 28 6.68 19 33.4 125 9.63 123 90.0 127 30.3 97 55.2 97 36.8 94
FlowNetS+ft+v [112]83.0 3.75 84 18.5 79 2.13 88 10.0 84 38.8 74 8.25 87 16.3 92 37.5 93 20.0 93 7.89 90 37.0 88 5.17 92 36.9 93 48.2 88 35.7 96 11.6 27 40.0 76 9.13 67 4.56 72 4.02 92 14.6 85 23.8 82 51.0 81 31.9 88
SuperFlow [81]83.1 3.40 74 16.6 68 1.81 81 15.3 94 41.5 83 15.9 94 17.0 95 35.2 85 27.6 98 10.0 101 43.1 94 8.60 100 36.1 90 44.3 65 46.7 112 13.1 42 39.8 75 11.3 87 2.49 43 4.24 96 4.26 28 30.0 95 54.7 95 40.8 100
Learning Flow [11]84.5 3.56 83 18.2 78 1.56 70 8.71 68 41.5 83 6.17 68 14.5 88 37.8 94 11.8 87 7.92 91 41.1 93 5.02 90 40.9 107 51.7 104 42.4 103 15.4 82 47.2 103 11.4 88 2.73 49 6.19 112 7.64 54 23.0 77 49.9 79 28.9 78
Second-order prior [8]84.6 3.49 77 18.6 80 1.79 78 9.83 79 40.6 81 7.83 82 14.1 85 39.0 95 9.82 80 6.20 81 31.4 76 3.83 82 34.7 86 49.4 97 27.6 78 18.7 96 52.1 111 11.4 88 9.18 109 3.60 88 20.1 101 19.1 61 48.2 76 24.4 64
Ad-TV-NDC [36]86.2 10.3 113 20.2 88 18.1 122 38.1 112 53.0 106 43.3 114 28.0 108 45.6 104 35.7 105 20.6 108 48.9 106 23.8 109 29.1 52 42.5 56 19.1 37 13.7 56 36.1 54 9.46 73 2.03 37 1.43 36 4.38 30 41.4 114 65.3 112 57.0 116
CNN-flow-warp+ref [117]86.6 4.85 99 24.6 101 2.62 100 13.2 91 41.8 85 12.9 91 17.7 97 39.6 97 25.0 96 8.67 95 44.8 98 5.78 93 38.1 98 48.2 88 43.5 106 13.7 56 40.4 82 9.32 70 1.80 32 1.29 35 9.16 61 33.2 103 57.7 103 42.9 101
StereoOF-V1MT [119]88.2 4.08 88 23.0 94 1.49 63 10.4 86 53.3 107 4.35 39 16.3 92 49.5 109 5.71 61 7.37 88 48.5 104 4.03 84 46.7 111 62.9 114 42.9 105 21.9 106 64.7 117 17.0 101 1.58 26 1.87 51 9.43 63 33.2 103 66.2 113 36.5 93
BlockOverlap [61]90.7 4.14 92 17.3 73 3.50 103 23.3 101 43.1 93 25.5 103 21.0 100 37.2 90 27.8 99 13.1 102 39.0 91 13.7 104 28.8 48 38.6 36 28.2 80 18.7 96 37.2 58 13.3 95 12.6 118 6.40 115 40.8 119 26.8 91 45.8 67 43.3 103
Shiralkar [42]92.2 4.11 90 23.3 95 1.52 66 8.88 71 44.5 96 5.07 49 14.5 88 41.9 100 6.96 67 7.32 87 44.7 97 4.37 87 38.8 103 55.2 109 33.1 88 26.7 113 60.7 112 18.5 106 10.4 112 3.38 87 32.9 113 26.7 90 61.6 107 29.5 81
SegOF [10]92.3 6.07 107 25.2 106 3.62 105 36.7 110 55.3 108 41.7 112 26.1 104 43.8 102 39.2 108 15.2 104 45.4 99 12.3 102 46.5 110 56.0 110 57.5 116 18.2 94 49.9 106 14.9 98 0.19 4 0.71 23 0.86 7 31.1 100 52.9 88 35.9 92
StereoFlow [44]93.1 28.4 126 55.1 127 37.7 126 81.1 127 92.6 127 77.8 126 65.0 125 82.9 127 51.1 121 69.6 127 90.7 127 65.5 124 52.7 118 67.5 118 44.9 110 8.13 4 33.5 43 6.60 16 0.05 2 0.37 16 0.17 2 32.2 101 54.8 96 40.4 99
HBpMotionGpu [43]93.8 5.00 100 21.6 91 2.81 102 31.1 107 49.5 101 35.0 107 26.3 106 45.3 103 37.1 107 9.18 98 39.3 92 7.65 99 33.7 78 45.6 77 32.2 86 15.7 84 37.4 61 9.89 78 5.71 90 4.19 95 12.5 76 33.4 105 56.3 99 47.8 108
SPSA-learn [13]94.5 5.52 104 25.3 108 4.12 107 25.2 104 50.0 102 26.8 104 25.1 103 45.7 105 36.7 106 19.0 105 54.2 108 20.8 106 38.6 100 48.4 90 44.4 108 17.9 92 45.3 99 17.6 103 1.60 27 0.54 20 5.27 38 39.6 111 57.9 104 53.3 113
2bit-BM-tele [98]95.3 5.17 102 23.4 96 3.54 104 16.3 95 40.3 78 16.7 96 15.2 90 34.6 83 14.2 91 8.49 94 37.6 90 6.75 97 32.8 73 44.5 68 28.6 83 24.3 109 43.9 94 22.5 112 15.6 123 8.72 121 50.0 124 26.2 88 50.8 80 37.5 96
FlowNet2 [122]95.7 7.84 109 30.7 109 2.58 99 41.4 114 65.2 114 44.4 115 29.6 110 48.0 107 46.8 116 5.31 77 24.0 46 3.06 78 47.8 112 64.8 117 36.3 99 17.8 91 44.3 97 13.4 96 2.93 54 8.71 120 5.22 36 30.2 96 61.7 108 28.8 77
IAOF2 [51]96.8 4.13 91 20.4 90 2.02 85 18.0 97 45.9 97 18.0 97 17.1 96 37.3 91 21.4 94 46.4 118 57.8 110 56.1 121 37.4 96 48.8 92 35.9 98 25.5 110 42.7 90 20.4 109 6.62 96 3.04 84 15.3 92 26.8 91 51.0 81 37.9 97
Black & Anandan [4]97.0 5.52 104 25.2 106 4.71 109 24.4 103 52.8 104 24.4 102 26.8 107 48.3 108 34.4 103 20.9 109 60.4 111 22.4 108 38.7 102 49.7 98 42.8 104 18.9 98 49.4 104 17.1 102 1.78 31 2.57 77 3.30 20 36.0 107 57.3 102 49.5 109
Filter Flow [19]97.4 5.13 101 23.5 97 2.40 97 20.5 99 51.3 103 19.6 98 23.3 102 42.6 101 35.3 104 27.2 111 48.8 105 28.3 111 39.4 104 49.1 93 44.6 109 17.9 92 40.5 83 11.8 91 7.39 101 7.67 118 11.5 71 26.6 89 46.0 68 33.9 89
Modified CLG [34]99.2 7.42 108 31.9 111 5.50 110 31.7 108 52.9 105 37.6 110 28.4 109 50.8 110 40.4 110 20.3 107 60.5 112 21.3 107 39.5 106 51.2 101 42.2 102 14.2 66 45.8 102 11.5 90 3.24 57 1.70 49 9.31 62 40.3 112 65.1 111 54.7 115
IAOF [50]100.3 5.70 106 24.0 99 3.65 106 30.0 106 48.7 99 33.9 106 26.2 105 47.8 106 29.5 100 28.0 112 51.3 107 32.9 112 37.3 95 50.2 99 34.4 93 26.2 111 50.9 109 18.0 105 5.85 92 1.63 42 11.7 72 36.3 108 59.4 106 51.9 111
GraphCuts [14]102.1 5.45 103 24.7 102 2.64 101 24.3 102 55.8 110 21.9 101 21.4 101 40.8 99 32.4 102 9.25 100 46.4 100 6.31 95 38.6 100 51.7 104 33.8 91 28.8 116 39.7 74 18.7 107 12.1 117 2.87 81 35.1 116 38.5 109 58.4 105 53.9 114
2D-CLG [1]102.6 14.0 117 40.7 118 8.09 115 45.8 116 59.5 111 54.5 119 36.8 116 60.4 117 47.3 117 48.9 120 75.1 121 54.2 120 44.9 108 54.8 107 52.5 113 19.0 99 50.3 108 17.8 104 1.26 21 0.07 6 4.43 31 47.4 119 71.2 119 59.9 120
GroupFlow [9]102.9 8.95 112 33.2 112 7.07 112 43.6 115 70.7 119 45.5 116 32.7 112 59.8 116 42.4 113 13.2 103 46.6 101 12.4 103 51.1 115 70.0 120 34.2 92 30.8 118 62.8 114 33.8 119 1.54 25 2.56 76 4.14 25 39.0 110 67.0 115 47.4 107
UnFlow [129]104.2 22.1 124 43.6 121 7.44 114 52.6 120 73.8 123 54.3 118 47.7 122 74.9 125 47.5 118 26.4 110 68.7 116 24.7 110 64.5 123 75.5 124 65.6 123 28.5 115 67.0 121 27.1 116 0.19 4 1.87 51 0.05 1 30.4 98 62.4 109 36.8 94
Nguyen [33]105.0 8.19 110 31.4 110 4.40 108 54.9 121 55.7 109 70.2 122 33.6 114 54.6 111 43.3 114 43.5 117 60.9 113 50.5 118 45.1 109 54.9 108 54.2 114 21.1 103 49.4 104 21.0 110 2.92 53 1.87 51 7.39 52 44.7 117 66.2 113 58.5 118
Horn & Schunck [3]106.3 8.56 111 35.5 113 7.11 113 29.4 105 65.4 115 28.1 105 33.4 113 64.4 119 41.2 112 30.6 113 67.2 115 33.6 113 49.1 114 61.3 111 55.5 115 26.4 112 64.7 117 26.1 115 3.02 55 3.95 90 2.44 15 48.8 120 75.0 121 58.8 119
SILK [79]108.2 10.7 114 35.7 114 14.6 119 37.7 111 64.2 112 42.5 113 32.3 111 59.3 114 40.6 111 19.9 106 56.8 109 20.4 105 51.6 117 62.6 113 59.6 119 27.2 114 63.0 115 23.2 113 4.92 81 1.68 48 13.1 78 46.9 118 71.6 120 61.7 122
TI-DOFE [24]111.7 21.2 122 43.7 122 34.5 125 64.6 125 71.9 122 76.5 125 47.1 121 75.6 126 53.7 122 57.3 122 76.4 122 65.6 125 51.5 116 63.9 115 60.0 120 30.4 117 65.8 120 33.2 117 2.24 40 1.65 45 5.22 36 59.1 124 82.1 126 71.2 125
Periodicity [78]112.0 11.0 115 41.5 119 5.85 111 35.3 109 64.5 113 36.8 109 51.0 123 55.5 112 61.7 124 49.6 121 81.4 124 48.4 117 66.0 125 83.6 127 59.0 117 46.1 123 76.4 125 43.0 123 1.67 30 5.33 106 8.13 56 48.8 120 78.9 123 57.5 117
Heeger++ [104]113.0 24.4 125 45.8 123 9.72 117 47.8 118 85.9 126 37.8 111 66.2 126 69.8 122 75.2 125 59.7 124 88.6 126 56.9 122 66.6 126 80.3 126 62.4 122 65.6 127 87.3 127 67.1 127 4.14 69 1.26 34 7.39 52 41.9 115 68.8 116 43.4 105
SLK [47]115.6 17.8 120 50.1 126 21.7 123 62.2 124 77.8 125 74.7 124 40.4 119 72.7 124 48.1 119 66.2 125 73.9 119 76.1 127 60.9 121 71.2 121 72.9 127 33.1 119 68.3 123 35.4 121 5.99 93 1.09 28 11.4 70 60.5 126 81.9 125 75.0 126
FFV1MT [106]116.1 22.0 123 42.2 120 9.20 116 41.0 113 76.9 124 36.1 108 66.5 127 71.9 123 79.2 126 58.5 123 87.7 125 57.1 123 64.8 124 76.5 125 70.3 126 64.8 126 85.8 126 65.3 126 4.70 76 4.86 101 11.3 69 41.9 115 68.8 116 43.4 105
Adaptive flow [45]117.1 16.0 118 36.3 115 17.5 121 57.9 122 67.3 116 64.2 121 38.7 118 59.2 113 48.8 120 39.9 115 69.2 117 44.2 116 49.0 113 62.0 112 43.6 107 39.1 122 62.2 113 34.3 120 34.2 126 23.4 126 82.8 125 40.6 113 64.9 110 51.9 111
FOLKI [16]118.3 13.4 116 45.9 124 16.0 120 48.5 119 67.4 117 57.8 120 36.6 115 66.2 120 40.3 109 32.2 114 66.9 114 37.2 114 52.9 119 64.2 116 60.1 121 34.7 120 65.7 119 40.2 122 12.9 119 7.56 117 33.8 115 55.3 123 78.0 122 70.7 124
PGAM+LK [55]120.3 17.6 119 48.4 125 26.7 124 45.8 116 71.8 121 49.9 117 38.3 117 67.6 121 44.6 115 42.8 116 79.5 123 42.6 115 56.5 120 69.9 119 59.0 117 37.8 121 70.5 124 33.6 118 23.5 124 15.0 125 48.8 123 54.6 122 80.9 124 61.1 121
Pyramid LK [2]120.8 19.8 121 36.7 116 40.3 127 61.8 123 68.8 118 74.6 123 43.5 120 63.7 118 58.2 123 46.9 119 72.7 118 54.1 119 61.5 122 74.5 123 65.8 124 51.3 124 64.1 116 50.1 124 10.8 114 6.30 113 31.9 112 68.2 127 84.9 127 84.8 127
HCIC-L [99]123.5 29.6 127 37.7 117 11.4 118 77.3 126 71.7 120 90.9 127 63.0 124 59.4 115 83.6 127 66.9 126 74.8 120 69.8 126 68.1 127 74.1 122 66.1 125 54.4 125 67.4 122 52.5 125 58.9 127 43.5 127 89.4 126 59.4 125 70.0 118 67.4 123
AdaConv-v1 [126]128.0 82.5 128 78.4 128 94.0 128 98.6 128 99.3 128 97.9 128 99.9 128 99.9 128 99.9 128 97.4 128 96.9 128 99.7 128 100.0 128 99.9 128 99.8 128 93.4 128 95.5 128 93.4 128 86.5 128 85.7 128 99.4 128 99.9 128 99.9 128 99.9 128
SepConv-v1 [127]128.0 82.5 128 78.4 128 94.0 128 98.6 128 99.3 128 97.9 128 99.9 128 99.9 128 99.9 128 97.4 128 96.9 128 99.7 128 100.0 128 99.9 128 99.8 128 93.4 128 95.5 128 93.4 128 86.5 128 85.7 128 99.4 128 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.