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        
R5.0
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]10.6 7.69 2 26.2 5 3.54 2 7.19 14 30.7 27 6.11 20 5.88 5 19.3 8 4.53 14 4.01 10 23.9 16 2.00 12 11.7 1 16.4 3 5.52 3 10.4 14 29.7 9 9.30 10 5.25 15 20.0 23 2.61 7 1.88 7 6.47 27 0.19 1
MDP-Flow2 [68]11.8 10.3 20 30.4 19 6.57 20 5.28 2 23.9 3 4.13 3 5.46 3 17.6 3 3.58 7 4.49 18 25.3 21 2.11 15 15.8 21 21.4 21 10.2 26 10.6 15 29.9 10 9.87 16 4.44 7 19.3 16 2.81 8 1.39 2 4.82 5 1.11 3
OFLAF [77]12.5 8.96 9 27.9 9 4.57 8 7.36 17 26.4 8 6.68 22 4.80 2 14.9 2 3.37 6 4.43 17 21.2 8 2.81 26 13.4 7 19.0 10 7.08 7 11.4 24 28.0 3 9.28 9 5.35 17 19.1 15 3.15 10 2.47 16 5.73 18 5.50 31
NN-field [71]13.7 8.65 5 28.3 10 4.00 4 8.38 26 33.1 41 7.44 28 5.86 4 19.0 7 4.53 14 3.15 2 21.4 10 1.25 3 12.0 4 16.9 5 5.41 2 6.58 2 20.2 1 3.45 1 8.64 45 23.6 50 2.88 9 2.47 16 8.48 38 0.20 2
PMMST [114]19.3 12.9 46 32.6 31 8.45 42 10.2 42 30.5 24 10.7 57 7.39 15 22.2 12 6.31 31 3.87 9 13.5 1 2.73 23 14.0 10 18.7 9 7.52 8 10.9 21 28.9 6 9.95 17 4.99 12 20.8 27 3.18 11 1.52 3 3.87 2 1.12 4
ComponentFusion [96]21.6 8.86 8 28.7 13 5.91 14 6.30 5 24.2 4 5.98 15 6.79 10 21.6 10 4.99 20 4.11 11 24.4 17 2.04 14 16.2 29 22.0 26 11.3 37 13.4 40 40.4 56 12.4 55 7.66 37 21.3 33 5.22 32 2.05 8 5.21 9 3.61 15
HAST [109]22.1 7.13 1 21.8 1 3.37 1 7.28 15 26.1 7 6.10 19 3.86 1 12.2 1 0.97 1 3.85 8 21.3 9 1.50 4 11.7 1 16.5 4 4.64 1 15.1 61 35.5 32 14.0 66 19.4 85 36.3 93 39.0 113 1.24 1 3.55 1 1.32 5
ALD-Flow [66]22.2 8.44 4 27.6 8 4.09 5 6.49 7 27.2 12 5.04 9 7.66 20 24.1 22 3.72 9 4.58 20 27.1 32 2.01 13 16.0 24 22.7 34 8.55 12 9.39 8 33.3 21 8.46 8 7.21 34 18.5 10 17.3 83 4.06 46 11.1 52 5.94 39
nLayers [57]23.0 8.66 6 25.4 3 4.54 6 13.0 78 32.1 34 13.7 85 7.74 22 21.8 11 8.85 60 3.29 4 18.2 2 1.89 9 11.8 3 16.2 2 6.65 6 12.2 29 28.4 4 10.3 20 8.59 44 21.5 37 4.98 28 2.36 14 5.74 19 5.08 27
NNF-EAC [103]23.2 10.9 28 31.8 26 6.97 21 6.64 9 26.4 8 5.65 11 6.61 9 20.7 9 4.44 13 5.48 39 27.1 32 2.97 31 16.5 32 22.3 30 11.1 34 12.1 28 30.7 11 9.97 18 6.42 25 21.4 35 3.94 22 2.95 29 7.51 33 4.63 24
LME [70]23.4 9.71 15 29.0 15 6.46 19 5.49 4 22.8 2 4.79 7 8.62 36 22.4 13 11.2 71 4.73 21 28.3 46 2.35 18 16.5 32 21.9 24 12.6 50 10.7 16 34.0 25 9.81 15 5.57 18 21.3 33 3.87 20 2.40 15 6.32 24 4.52 22
TC/T-Flow [76]24.2 9.15 10 32.1 28 3.69 3 6.89 12 31.2 30 4.32 4 7.32 14 23.1 17 4.08 10 5.14 30 27.2 35 2.80 25 15.6 20 21.9 24 9.71 20 8.63 6 29.2 7 8.40 6 6.72 27 20.2 24 19.7 97 3.86 44 9.43 42 6.28 45
WLIF-Flow [93]25.3 9.40 14 27.1 6 6.06 16 10.0 39 33.0 39 9.71 42 7.26 13 22.4 13 5.90 30 4.53 19 23.7 14 2.56 22 16.1 25 22.3 30 11.3 37 12.8 32 33.2 20 10.4 23 6.85 29 18.8 13 7.36 44 2.80 26 6.48 28 5.62 34
FC-2Layers-FF [74]26.2 9.97 16 28.7 13 7.96 31 10.4 44 35.3 49 9.95 43 6.11 7 18.1 6 6.82 35 4.12 12 20.9 6 2.48 21 13.3 6 17.7 6 9.47 17 13.5 41 32.6 15 11.3 34 14.0 69 26.0 63 12.5 64 1.84 5 4.18 3 4.53 23
RNLOD-Flow [121]26.5 7.93 3 24.8 2 5.41 12 8.33 25 31.7 31 6.84 24 7.47 16 23.3 18 4.60 16 3.62 6 21.1 7 1.66 7 14.4 12 21.0 15 7.80 9 12.7 31 32.9 16 12.2 54 17.4 81 34.4 88 20.7 99 2.55 19 5.57 16 5.30 29
Layers++ [37]27.2 10.2 18 29.1 16 8.58 45 10.8 49 30.6 25 11.0 63 5.90 6 17.7 4 6.34 32 3.40 5 18.2 2 1.66 7 12.2 5 16.0 1 9.69 19 13.9 46 33.6 22 11.9 48 13.9 68 27.9 68 8.74 49 2.33 13 4.94 6 5.70 37
FESL [72]28.4 8.67 7 25.6 4 4.73 9 13.6 83 39.4 84 12.9 80 8.22 30 24.3 26 7.10 39 3.76 7 20.3 5 2.12 16 14.1 11 20.1 12 9.08 14 11.2 23 31.0 12 9.80 14 11.3 54 30.4 76 7.66 46 2.10 9 5.40 12 2.40 8
SVFilterOh [111]28.8 12.7 44 28.4 11 8.58 45 9.41 34 29.1 19 8.31 33 6.39 8 17.7 4 5.05 22 3.23 3 18.8 4 0.95 2 14.9 15 20.9 14 6.51 5 13.2 36 32.1 14 11.6 41 22.0 91 46.5 116 30.5 108 1.61 4 4.97 8 2.45 10
TC-Flow [46]29.8 9.24 11 30.9 22 5.24 11 5.48 3 25.7 6 4.01 2 7.25 12 23.3 18 2.66 3 5.52 40 28.4 47 3.26 37 16.7 35 23.9 40 9.52 18 11.7 25 36.8 37 11.5 40 6.69 26 21.4 35 19.0 94 4.21 48 10.6 49 6.91 57
OAR-Flow [125]31.8 10.8 26 34.0 40 6.23 17 8.94 29 31.9 32 7.53 29 10.4 51 30.3 53 7.58 48 5.60 44 25.9 27 3.04 33 17.0 36 24.0 41 9.35 15 8.62 5 33.0 17 7.87 4 3.37 4 14.6 3 5.54 35 4.80 55 10.4 48 9.44 72
Efficient-NL [60]32.4 9.31 12 27.5 7 5.65 13 12.1 70 38.1 69 11.2 65 8.07 29 24.1 22 6.69 34 5.39 36 25.9 27 3.51 47 14.9 15 21.1 17 9.39 16 14.0 49 35.2 30 11.1 31 11.5 55 25.7 61 6.90 41 2.19 11 5.45 14 1.94 7
AGIF+OF [85]32.5 10.4 21 29.4 17 7.42 24 12.4 71 37.5 62 12.4 75 7.91 27 24.3 26 7.24 41 4.84 24 23.8 15 2.91 28 14.8 14 20.6 13 9.72 21 13.2 36 36.5 35 10.5 25 7.06 31 20.8 27 7.54 45 3.02 34 6.27 22 6.37 47
IROF++ [58]33.0 10.2 18 30.9 22 7.02 22 11.1 55 38.1 69 10.7 57 8.32 34 25.1 31 7.61 49 5.83 47 28.0 43 4.08 56 15.4 19 21.3 19 9.83 22 13.6 42 38.0 46 11.3 34 5.83 19 20.8 27 1.97 6 2.32 12 5.71 17 4.87 26
PMF [73]33.5 11.6 37 29.9 18 4.55 7 7.81 20 30.2 21 6.00 16 7.17 11 23.3 18 3.21 5 4.88 26 23.1 12 2.42 19 13.6 8 18.5 7 6.49 4 16.1 69 42.7 66 15.4 75 27.2 105 43.5 113 28.9 105 2.15 10 4.96 7 4.81 25
PH-Flow [101]34.1 10.9 28 32.6 31 7.94 30 10.9 51 37.4 60 10.5 51 7.56 19 22.8 16 7.74 52 5.75 46 27.2 35 4.04 54 14.4 12 19.8 11 8.81 13 13.0 34 33.6 22 11.1 31 12.7 62 23.1 46 17.6 85 1.84 5 4.19 4 4.50 21
Classic+CPF [83]35.2 10.9 28 31.7 25 7.88 29 11.5 61 37.9 66 10.9 61 8.27 32 25.1 31 7.51 47 5.05 28 25.8 26 3.16 35 15.1 17 21.0 15 10.6 27 13.1 35 34.6 28 10.3 20 9.87 49 22.0 42 13.1 70 2.68 22 5.85 20 5.53 32
COFM [59]35.6 10.1 17 32.0 27 7.63 25 8.06 22 30.4 23 7.17 25 8.93 40 25.9 37 8.04 56 4.17 13 24.9 19 1.63 6 18.8 48 24.0 41 18.6 87 14.4 53 33.0 17 11.7 43 8.15 40 20.4 25 14.7 77 3.16 36 5.36 11 8.09 67
Ramp [62]36.5 10.9 28 32.7 35 7.96 31 10.9 51 37.1 57 10.6 53 7.85 23 24.2 24 7.41 45 5.29 33 27.0 31 3.44 42 16.1 25 22.3 30 10.8 28 13.8 45 35.4 31 11.0 30 11.6 56 21.1 31 18.2 89 2.52 18 5.44 13 5.23 28
Sparse-NonSparse [56]37.1 10.7 24 32.5 30 8.38 40 10.9 51 36.8 55 10.7 57 7.95 28 24.5 30 7.30 43 5.42 38 27.6 38 3.49 46 16.1 25 22.1 27 11.0 32 13.3 38 36.0 34 10.6 26 10.6 52 21.1 31 10.9 55 2.91 27 5.93 21 6.18 43
Correlation Flow [75]37.2 11.9 41 35.3 41 6.03 15 6.85 11 28.0 14 4.77 6 8.29 33 25.8 35 2.17 2 4.84 24 27.2 35 2.77 24 18.5 46 25.9 53 11.7 39 16.9 76 39.5 52 16.7 81 12.1 59 24.6 57 17.8 86 2.59 20 7.33 32 3.08 11
LSM [39]37.6 10.4 21 32.6 31 8.24 35 10.8 49 37.4 60 10.4 50 7.85 23 24.3 26 7.05 37 5.32 34 27.6 38 3.41 41 15.8 21 21.5 22 11.1 34 13.7 43 35.6 33 10.9 29 13.0 63 23.2 47 12.5 64 2.99 33 6.43 26 6.14 42
ProbFlowFields [128]38.5 16.2 62 47.8 74 11.7 71 8.96 30 31.0 28 8.86 36 9.73 46 28.4 45 10.1 65 6.09 54 25.5 23 4.53 65 18.2 44 25.5 47 10.9 30 9.76 12 34.2 26 11.7 43 4.63 8 18.8 13 3.79 19 2.95 29 8.94 41 3.52 13
Classic+NL [31]40.0 10.5 23 31.4 24 8.38 40 11.1 55 37.9 66 10.6 53 7.87 25 24.0 21 7.48 46 5.57 42 27.6 38 3.62 49 15.8 21 21.5 22 10.8 28 14.1 50 37.4 41 11.4 37 14.8 72 25.9 62 13.4 73 2.61 21 5.29 10 6.10 41
FMOF [94]41.2 11.0 35 30.4 19 8.33 39 13.0 78 38.5 75 12.6 76 7.51 17 22.6 15 7.34 44 5.06 29 25.2 20 3.44 42 15.3 18 21.3 19 9.87 24 14.9 60 33.1 19 11.4 37 11.7 58 24.3 56 15.0 79 3.92 45 8.59 39 6.28 45
S2D-Matching [84]41.4 10.7 24 32.2 29 8.71 47 10.7 47 36.6 54 10.2 46 8.94 41 27.2 41 6.96 36 5.17 31 26.0 29 3.36 40 16.3 30 22.1 27 10.9 30 14.4 53 37.0 38 11.7 43 16.4 76 26.0 63 16.5 81 2.79 25 5.49 15 6.49 48
IROF-TV [53]44.2 11.6 37 35.3 41 9.03 50 11.2 58 38.2 72 10.9 61 8.85 39 26.5 38 7.73 51 6.04 53 33.0 70 3.62 49 17.1 37 23.1 36 13.5 59 16.3 70 44.8 73 13.5 64 3.41 5 16.9 5 1.13 3 2.71 23 6.80 30 5.67 36
TV-L1-MCT [64]44.4 10.9 28 30.5 21 8.56 44 13.8 86 40.9 93 13.2 82 8.68 37 25.8 35 7.98 55 4.83 23 25.7 24 3.26 37 17.4 39 23.5 39 13.7 63 14.8 59 36.7 36 12.7 58 5.84 20 19.4 17 10.1 53 3.53 40 6.42 25 6.63 51
Aniso-Texture [82]45.6 9.33 13 28.5 12 7.26 23 9.17 32 26.8 11 10.2 46 10.1 50 29.2 48 7.28 42 2.77 1 22.6 11 0.94 1 19.9 61 27.1 65 13.3 57 14.6 56 38.5 49 12.4 55 31.5 111 46.3 115 18.2 89 4.45 51 10.1 45 6.60 50
2DHMM-SAS [92]45.8 10.9 28 32.6 31 8.06 33 11.5 61 39.5 85 10.6 53 10.0 49 28.3 44 7.91 54 5.93 49 28.2 45 4.07 55 16.1 25 22.3 30 11.0 32 13.7 43 38.3 48 11.1 31 12.3 61 23.2 47 18.0 88 3.08 35 6.48 28 6.24 44
AggregFlow [97]45.9 13.9 49 33.8 38 11.2 66 13.7 84 39.6 87 12.6 76 12.0 64 31.3 54 13.7 80 5.40 37 23.5 13 3.44 42 17.5 40 25.4 45 7.98 11 8.57 4 25.9 2 8.42 7 7.00 30 24.1 55 4.53 25 5.53 65 9.80 44 11.7 83
CostFilter [40]46.1 14.1 50 36.2 47 8.48 43 8.61 28 30.6 25 7.43 27 8.26 31 26.9 40 4.40 12 5.72 45 28.1 44 3.24 36 13.7 9 18.5 7 7.81 10 16.6 73 45.0 78 16.0 78 26.8 103 48.6 119 32.7 109 2.93 28 7.59 34 5.38 30
SimpleFlow [49]46.2 11.6 37 33.7 37 8.98 49 12.5 74 38.9 78 12.6 76 10.4 51 29.3 49 9.20 61 5.99 50 27.6 38 4.08 56 16.3 30 22.2 29 11.1 34 16.7 74 37.4 41 12.7 58 8.29 41 19.9 21 6.11 39 2.74 24 6.28 23 5.86 38
Adaptive [20]46.7 10.9 28 33.8 38 4.92 10 10.5 45 35.0 48 9.53 41 12.2 65 33.7 58 7.68 50 5.57 42 30.3 56 2.95 30 21.7 84 26.7 61 20.6 93 10.8 18 34.9 29 7.26 3 14.0 69 28.8 71 4.88 26 4.50 53 10.2 47 6.84 55
MDP-Flow [26]47.2 12.2 43 40.6 54 8.88 48 9.32 33 28.3 16 10.5 51 9.09 42 28.1 43 9.37 63 6.03 52 30.6 58 3.99 52 17.2 38 23.1 36 12.4 45 13.9 46 42.7 66 12.5 57 7.10 33 23.6 50 4.09 24 5.35 61 13.2 61 7.09 60
RFlow [90]47.5 14.8 51 43.9 63 11.2 66 6.64 9 26.6 10 5.76 12 11.7 60 35.9 67 5.04 21 4.31 15 27.1 32 1.94 10 19.4 52 26.8 63 13.0 56 14.7 57 42.2 63 11.8 47 13.1 64 22.2 43 13.1 70 5.87 70 14.1 68 8.71 70
Occlusion-TV-L1 [63]48.2 12.9 46 36.1 46 8.26 38 9.51 37 32.7 36 8.99 37 12.3 66 34.4 62 8.27 57 5.53 41 29.8 54 3.04 33 20.5 72 28.5 83 13.8 65 9.95 13 37.9 43 11.6 41 7.64 36 21.8 41 3.47 13 5.69 67 13.9 66 7.59 64
OFH [38]49.5 15.0 53 40.9 55 14.4 82 7.06 13 29.9 20 5.37 10 10.8 54 33.1 57 4.86 19 5.84 48 30.6 58 3.46 45 19.5 54 26.1 54 15.3 71 15.6 65 46.5 85 16.6 80 4.19 6 21.7 39 3.74 18 5.39 63 15.4 77 7.23 61
MLDP_OF [89]51.9 18.8 82 51.3 89 16.0 85 8.16 23 32.0 33 6.76 23 10.7 53 31.9 56 5.45 25 4.81 22 26.1 30 2.44 20 18.7 47 24.3 43 13.7 63 15.6 65 37.9 43 18.6 89 19.2 84 28.5 69 38.7 112 3.53 40 7.25 31 4.27 19
DeepFlow2 [108]53.8 15.0 53 43.6 62 11.0 62 10.1 40 34.2 45 9.29 39 12.9 68 36.8 68 11.1 70 7.47 69 32.1 65 4.75 69 17.8 41 25.4 45 9.97 25 10.7 16 40.2 55 10.3 20 6.78 28 18.7 12 13.3 72 9.05 89 17.3 85 15.3 93
Steered-L1 [118]54.5 11.4 36 37.9 49 7.71 28 4.42 1 21.7 1 3.76 1 7.71 21 25.7 33 4.29 11 4.91 27 29.8 54 2.26 17 20.2 67 26.7 61 16.6 78 18.1 81 46.1 84 14.6 68 32.4 112 37.9 100 51.5 122 8.58 86 15.5 79 15.2 92
S2F-IF [123]55.6 18.0 76 51.9 92 10.9 59 11.1 55 38.6 76 10.6 53 13.9 73 40.6 82 13.4 79 7.68 76 32.6 67 5.18 75 19.7 59 27.2 67 13.3 57 10.8 18 39.5 52 11.9 48 4.99 12 19.9 21 6.26 40 3.26 38 10.1 45 3.57 14
Classic++ [32]55.8 10.8 26 32.7 35 8.25 37 10.5 45 32.9 37 10.7 57 10.8 54 31.6 55 8.46 58 5.25 32 29.7 53 2.99 32 20.0 63 28.0 78 13.9 66 15.2 63 44.1 71 11.9 48 17.3 80 26.2 65 18.3 91 5.82 69 12.7 57 8.14 68
Sparse Occlusion [54]55.8 12.7 44 35.8 45 8.24 35 12.4 71 33.4 42 13.4 83 9.67 44 29.1 47 6.55 33 5.99 50 28.5 48 3.56 48 19.4 52 26.4 59 12.4 45 14.7 57 39.4 51 11.7 43 37.7 119 48.6 119 17.8 86 3.66 42 9.43 42 5.64 35
PGM-C [120]55.9 17.7 71 50.5 83 11.0 62 11.9 65 39.1 80 11.6 69 13.9 73 40.4 81 13.3 78 7.52 72 35.8 82 4.62 68 19.6 57 27.5 70 12.4 45 9.48 10 37.9 43 9.36 11 4.63 8 16.9 5 5.02 29 4.83 56 14.2 71 6.69 52
CPM-Flow [116]56.6 17.7 71 50.5 83 11.0 62 11.9 65 39.0 79 11.7 70 13.7 72 39.8 79 13.2 76 7.49 71 35.5 80 4.58 67 19.5 54 27.2 67 12.3 44 9.44 9 37.3 40 9.46 12 5.05 14 19.5 18 5.17 31 5.21 59 14.8 73 7.36 62
BriefMatch [124]56.9 11.8 40 35.7 44 6.41 18 7.52 18 30.3 22 5.97 14 7.54 18 24.2 24 4.62 17 4.28 14 25.4 22 1.98 11 20.6 74 26.2 57 20.9 95 26.8 108 49.2 88 28.2 110 22.8 95 35.9 92 39.6 116 9.81 92 15.1 76 18.3 101
ACK-Prior [27]57.2 19.5 86 41.5 57 14.3 81 6.57 8 27.6 13 4.53 5 7.87 25 25.7 33 3.70 8 4.33 16 25.7 24 1.53 5 20.5 72 25.6 50 18.3 84 23.1 99 44.0 70 18.5 88 29.9 107 33.1 83 45.6 120 7.91 83 14.8 73 11.7 83
NL-TV-NCC [25]57.7 16.5 65 40.4 52 9.10 51 10.7 47 37.0 56 8.07 31 8.59 35 26.8 39 3.17 4 6.24 56 33.4 71 3.26 37 21.4 81 29.7 95 12.7 52 21.2 91 48.2 87 17.3 84 13.4 67 35.6 91 13.0 69 4.73 54 12.8 58 3.24 12
EpicFlow [102]57.8 17.7 71 50.6 85 10.9 59 12.0 68 39.3 83 11.7 70 14.5 76 42.2 86 13.2 76 7.47 69 35.5 80 4.57 66 19.8 60 27.6 71 12.8 54 9.73 11 38.1 47 10.1 19 4.63 8 17.2 7 4.88 26 5.31 60 14.3 72 7.47 63
CombBMOF [113]58.2 15.2 55 48.2 76 7.67 26 11.3 59 34.5 46 9.95 43 8.75 38 27.2 41 5.37 24 7.60 75 32.1 65 5.65 81 18.0 43 23.0 35 13.9 66 21.7 94 44.9 75 24.3 102 22.6 93 37.2 95 14.5 76 2.97 32 7.73 36 4.35 20
Complementary OF [21]59.2 20.9 88 51.7 90 21.5 94 6.41 6 28.3 16 4.86 8 9.56 43 30.2 52 5.62 26 8.21 80 31.4 61 6.20 84 19.2 50 25.6 50 15.5 72 21.5 93 49.3 89 17.4 85 6.34 23 19.8 20 11.5 58 6.44 76 16.1 82 10.2 76
FlowFields [110]59.6 18.3 77 51.9 92 11.1 65 11.9 65 39.5 85 11.5 67 14.8 77 43.3 89 14.2 81 7.96 78 33.5 74 5.52 78 19.9 61 27.6 71 13.6 61 11.0 22 40.5 57 12.1 52 4.93 11 19.7 19 5.34 34 3.85 43 12.3 56 3.89 16
TF+OM [100]60.4 14.8 51 35.4 43 7.68 27 9.06 31 28.4 18 9.32 40 11.6 59 28.4 45 16.0 87 6.43 57 29.0 51 4.29 59 20.2 67 25.6 50 18.4 85 17.9 79 38.5 49 16.9 83 16.6 77 33.8 84 14.7 77 6.87 79 15.5 79 9.68 73
ROF-ND [107]61.0 18.4 78 45.8 70 11.5 69 7.31 16 25.4 5 6.02 17 9.70 45 29.4 50 4.66 18 9.09 86 28.7 49 5.98 83 21.6 83 29.5 92 14.5 69 19.9 87 44.8 73 15.3 74 33.3 116 41.0 105 30.1 107 2.95 29 7.63 35 2.41 9
TV-L1-improved [17]62.4 11.9 41 36.8 48 8.23 34 8.49 27 31.0 28 7.83 30 11.9 62 33.7 58 7.19 40 5.35 35 28.9 50 2.91 28 20.3 70 28.0 78 12.0 42 27.2 110 55.4 105 30.4 112 23.1 98 38.0 101 22.9 102 5.61 66 14.0 67 7.74 66
ComplOF-FED-GPU [35]62.8 17.9 74 52.0 94 15.4 84 7.90 21 33.9 44 5.82 13 10.8 54 34.2 60 5.67 27 6.99 64 31.5 62 4.51 64 19.2 50 26.3 58 12.9 55 18.2 82 50.5 95 18.6 89 15.1 73 23.6 50 22.3 101 5.37 62 15.4 77 6.76 53
DeepFlow [86]62.8 17.5 70 46.9 72 16.5 86 11.8 63 35.8 50 11.2 65 15.1 79 39.6 78 15.2 85 7.81 77 32.6 67 5.12 74 17.8 41 25.5 47 9.86 23 12.0 27 44.9 75 11.4 37 6.11 22 18.0 8 12.8 67 10.8 98 18.7 92 18.8 103
TCOF [69]62.8 17.2 69 45.4 69 15.3 83 12.6 75 37.6 63 12.3 73 15.7 81 39.5 76 16.6 88 6.72 60 27.7 42 4.48 63 22.5 90 30.9 104 11.9 40 9.21 7 28.4 4 10.8 28 22.9 96 35.0 90 9.29 50 4.22 49 11.3 53 6.79 54
EPPM w/o HM [88]63.7 19.4 84 53.2 96 11.2 66 8.23 24 34.8 47 6.07 18 11.1 57 35.1 65 5.89 29 7.31 67 33.4 71 4.76 70 18.9 49 23.2 38 17.1 80 21.3 92 50.3 94 20.1 93 20.7 89 30.3 75 40.9 118 3.20 37 8.13 37 5.59 33
HBM-GC [105]64.4 31.9 99 41.2 56 25.6 99 13.2 80 32.9 37 14.2 87 9.93 48 24.4 29 8.75 59 10.1 94 24.7 18 6.95 92 16.6 34 21.1 17 13.6 61 18.5 83 33.7 24 15.5 76 33.9 118 47.5 118 20.1 98 3.38 39 8.62 40 5.97 40
Aniso. Huber-L1 [22]66.4 13.6 48 40.4 52 9.77 52 19.4 92 40.1 89 22.0 92 16.4 84 38.4 70 18.3 90 7.56 73 33.4 71 5.00 73 20.1 66 27.7 76 12.5 48 14.5 55 39.7 54 10.4 23 20.8 90 32.0 79 12.9 68 4.35 50 10.8 50 6.56 49
SIOF [67]66.7 16.5 65 40.1 51 10.8 58 10.3 43 37.1 57 9.10 38 16.4 84 38.3 69 18.4 91 8.56 81 35.1 78 5.87 82 21.3 80 28.5 83 16.5 77 17.6 78 43.6 69 19.7 92 7.08 32 21.6 38 3.65 16 6.65 77 16.1 82 10.9 80
F-TV-L1 [15]66.9 31.8 98 60.6 101 43.6 110 13.7 84 38.4 74 13.1 81 15.6 80 39.4 75 10.1 65 10.9 96 37.3 88 8.78 97 20.0 63 26.5 60 16.0 76 12.9 33 40.7 58 10.7 27 9.68 48 23.7 53 3.52 15 4.49 52 12.0 54 4.19 18
Rannacher [23]67.2 15.5 58 43.5 61 10.7 56 11.4 60 35.8 50 11.5 67 14.2 75 39.0 74 10.8 67 6.59 59 30.8 60 4.20 58 21.0 77 29.6 94 12.6 50 19.1 85 50.8 96 15.2 72 14.7 71 26.8 66 16.7 82 4.86 57 12.9 59 7.03 59
Brox et al. [5]69.7 18.5 79 51.2 87 20.8 93 14.0 87 37.8 65 15.1 89 13.6 71 38.8 72 11.7 72 7.20 66 36.8 85 4.02 53 23.0 94 28.5 83 24.3 103 10.8 18 45.3 80 9.57 13 7.81 38 22.7 44 1.58 4 9.61 91 19.2 95 15.0 91
LocallyOriented [52]69.8 15.8 61 41.5 57 10.9 59 15.0 89 44.5 98 13.7 85 17.6 88 43.4 90 14.2 81 7.16 65 31.5 62 4.82 71 21.0 77 29.0 88 12.5 48 11.7 25 34.5 27 12.9 60 11.6 56 29.6 72 12.0 61 7.94 84 18.4 89 11.1 81
SRR-TVOF-NL [91]70.0 22.3 94 44.7 65 12.5 75 12.0 68 38.1 69 10.2 46 14.8 77 40.6 82 10.9 69 6.13 55 34.1 75 2.81 26 19.6 57 25.5 47 13.5 59 16.4 71 42.4 64 13.0 61 30.5 108 42.5 109 18.3 91 6.41 75 11.0 51 12.0 85
CRTflow [80]70.7 16.5 65 49.5 79 10.6 55 9.63 38 33.8 43 8.65 35 13.1 69 38.8 72 7.80 53 6.86 62 34.3 76 4.44 62 20.0 63 27.8 77 12.2 43 31.4 115 59.0 111 36.7 116 10.3 50 30.4 76 12.0 61 8.56 85 20.4 101 12.9 89
DPOF [18]70.9 20.5 87 50.2 82 10.5 53 12.6 75 41.8 95 11.0 63 11.8 61 34.3 61 10.8 67 8.61 83 38.9 94 5.43 76 19.5 54 26.1 54 15.1 70 16.8 75 41.5 62 15.2 72 23.3 99 23.9 54 50.1 121 5.05 58 14.1 68 4.13 17
Bartels [41]72.6 19.3 83 39.6 50 22.4 96 9.47 36 28.2 15 10.0 45 9.91 47 29.7 51 7.09 38 9.18 88 29.3 52 7.40 94 21.7 84 27.6 71 21.1 96 19.1 85 44.4 72 24.2 101 23.0 97 36.3 93 36.2 110 7.46 82 14.9 75 11.5 82
Dynamic MRF [7]73.4 22.0 92 52.3 95 25.2 98 7.67 19 33.0 39 6.18 21 12.4 67 39.8 79 5.34 23 6.49 58 35.4 79 3.86 51 22.9 92 29.2 90 20.7 94 22.2 96 57.8 108 22.9 98 7.42 35 18.1 9 25.1 103 13.2 104 21.3 105 20.5 106
DF-Auto [115]75.6 19.4 84 46.0 71 10.5 53 26.6 101 46.1 100 31.1 101 23.7 99 46.1 93 37.0 101 9.05 85 36.8 85 5.59 79 21.7 84 29.1 89 17.4 82 7.80 3 31.8 13 7.93 5 19.5 86 37.4 97 3.25 12 10.9 99 19.6 97 16.4 95
TriangleFlow [30]75.8 18.7 81 43.9 63 18.0 87 10.1 40 37.2 59 8.18 32 11.9 62 35.5 66 5.81 28 6.72 60 34.6 77 4.37 61 26.7 110 34.7 111 23.4 100 23.1 99 49.6 90 23.5 99 16.7 78 37.2 95 16.3 80 6.85 78 17.3 85 10.3 77
CBF [12]75.8 15.2 55 44.8 66 12.1 73 23.7 97 37.7 64 30.9 100 13.2 70 34.6 63 14.5 83 6.86 62 32.8 69 4.32 60 22.6 91 28.4 82 20.2 91 15.6 65 41.0 60 12.1 52 32.9 114 39.7 103 29.8 106 5.49 64 13.2 61 8.30 69
Local-TV-L1 [65]76.0 24.6 95 51.2 87 30.0 100 22.5 96 40.6 91 25.2 95 23.5 98 46.1 93 28.3 96 9.73 91 37.4 89 6.92 91 18.3 45 25.2 44 12.7 52 13.9 46 43.2 68 12.0 51 5.25 15 20.6 26 5.15 30 15.8 109 21.0 103 32.1 114
SuperFlow [81]77.3 16.2 62 42.7 60 13.0 77 20.9 93 39.6 87 25.0 94 19.7 93 40.6 82 31.8 98 9.89 93 41.2 97 7.16 93 20.9 76 27.1 65 20.3 92 12.2 29 41.1 61 11.3 34 19.0 83 32.1 81 3.87 20 10.1 94 19.3 96 16.4 95
LDOF [28]77.4 17.1 68 48.0 75 12.9 76 13.3 81 40.6 91 12.2 72 15.8 82 42.4 87 12.7 75 9.70 90 44.0 99 6.27 86 20.7 75 28.0 78 16.8 79 14.3 52 45.9 83 13.8 65 8.36 42 23.3 49 7.98 47 11.2 101 21.2 104 18.3 101
CNN-flow-warp+ref [117]78.4 18.5 79 50.0 80 13.9 80 17.8 91 37.9 66 21.1 91 21.3 96 47.3 97 29.7 97 9.13 87 38.8 92 6.72 89 21.8 87 28.2 81 19.6 88 14.2 51 45.7 82 13.1 62 5.94 21 18.5 10 10.9 55 12.4 103 20.6 102 16.4 95
CLG-TV [48]78.6 15.7 60 42.2 59 11.7 71 20.9 93 39.2 82 24.8 93 16.4 84 39.5 76 18.0 89 9.23 89 37.9 90 6.54 87 22.9 92 30.0 99 17.9 83 16.5 72 47.2 86 14.2 67 19.9 87 30.4 76 11.5 58 5.79 68 14.1 68 6.98 58
TriFlow [95]79.2 21.3 90 44.9 67 13.5 79 16.0 90 36.5 53 18.7 90 18.2 90 38.6 71 27.8 95 7.35 68 30.3 56 5.59 79 21.2 79 27.3 69 18.5 86 15.1 61 37.1 39 15.0 71 49.5 124 41.7 107 95.6 127 6.38 74 13.3 63 9.77 74
p-harmonic [29]79.3 21.2 89 63.8 106 20.6 92 12.4 71 35.9 52 12.7 79 17.7 89 47.5 98 14.9 84 10.9 96 42.1 98 8.85 98 20.4 71 26.1 54 17.1 80 17.9 79 52.5 99 18.4 87 15.6 75 28.6 70 5.86 37 5.89 71 13.5 64 7.67 65
FlowNetS+ft+v [112]80.2 15.2 55 44.9 67 10.7 56 13.4 82 38.2 72 13.4 83 18.8 92 42.8 88 24.4 93 9.01 84 38.8 92 6.24 85 23.2 96 31.5 108 15.9 74 13.3 38 42.6 65 13.1 62 18.2 82 32.6 82 21.9 100 8.73 87 19.1 94 12.8 88
Second-order prior [8]82.5 15.6 59 48.2 76 12.1 73 12.6 75 39.1 80 12.3 73 16.2 83 44.6 91 12.2 74 7.57 74 31.6 64 5.45 77 22.2 88 30.6 101 14.3 68 20.8 90 56.8 107 17.7 86 28.0 106 33.8 84 27.1 104 7.43 81 17.4 87 10.4 79
StereoFlow [44]84.5 85.4 127 89.0 127 87.9 126 73.1 127 88.5 127 68.8 122 66.8 126 87.5 126 52.4 119 81.5 126 91.1 126 78.5 126 25.9 109 27.6 71 29.7 114 6.38 1 29.4 8 6.60 2 1.39 1 10.9 1 0.20 1 6.34 73 13.8 65 10.3 77
Fusion [6]85.5 17.9 74 57.7 98 18.6 88 9.42 35 32.3 35 10.2 46 11.4 58 34.8 64 11.7 72 8.57 82 40.2 95 6.89 90 25.0 106 30.8 103 24.9 107 23.9 102 52.3 98 25.0 105 33.3 116 43.4 111 19.3 96 9.01 88 18.8 93 13.4 90
Learning Flow [11]85.9 16.4 64 47.3 73 11.5 69 14.0 87 40.3 90 14.4 88 16.4 84 41.7 85 15.6 86 8.05 79 40.7 96 4.87 72 27.1 112 35.0 113 22.5 99 17.2 77 50.0 93 16.0 78 15.5 74 34.1 87 13.9 75 10.1 94 20.2 100 12.5 87
SegOF [10]88.0 28.8 97 51.1 86 13.2 78 37.3 110 51.8 110 44.6 112 30.0 104 53.0 102 43.3 109 27.0 109 49.6 104 22.4 105 24.0 102 27.6 71 28.4 113 24.9 105 58.5 109 24.4 103 2.04 2 16.2 4 0.47 2 10.0 93 16.5 84 16.7 98
Ad-TV-NDC [36]89.8 44.8 113 63.0 104 69.1 119 40.3 113 48.4 105 48.3 114 34.8 108 58.5 105 39.9 104 26.5 108 47.8 103 27.7 109 20.2 67 28.5 83 11.9 40 15.2 63 40.9 59 14.7 69 8.46 43 21.0 30 5.69 36 23.9 119 28.3 119 41.9 123
StereoOF-V1MT [119]90.3 21.7 91 68.0 111 20.3 91 11.8 63 50.4 108 7.18 26 20.7 95 62.8 108 9.21 62 9.80 92 50.8 105 6.56 88 27.9 114 35.8 114 23.9 101 25.0 106 67.3 116 24.0 100 8.00 39 27.7 67 12.2 63 13.4 105 23.5 109 15.8 94
Shiralkar [42]90.7 22.0 92 69.5 113 19.6 89 10.9 51 42.6 96 8.48 34 18.4 91 54.0 103 9.43 64 10.1 94 45.4 100 7.72 95 21.5 82 28.9 87 15.9 74 26.8 108 60.7 112 25.4 107 24.3 101 29.9 74 39.4 114 11.0 100 23.8 110 12.2 86
FlowNet2 [122]93.4 47.2 114 61.0 102 42.4 107 44.5 115 57.5 112 51.3 116 37.6 111 64.7 110 43.1 108 21.0 103 35.8 82 17.9 102 25.8 107 30.6 101 24.6 105 20.4 88 49.7 92 21.0 94 32.5 113 53.3 123 4.06 23 4.13 47 13.0 60 1.49 6
HBpMotionGpu [43]96.0 32.0 100 50.0 80 22.9 97 36.1 109 47.0 102 43.9 111 29.2 103 51.9 101 38.6 103 13.0 98 37.1 87 10.2 99 23.5 98 29.5 92 24.2 102 18.9 84 44.9 75 15.9 77 33.2 115 41.2 106 12.6 66 11.8 102 18.5 90 22.7 107
IAOF2 [51]96.2 25.3 96 49.2 78 22.2 95 24.6 98 44.3 97 28.6 99 20.0 94 45.4 92 25.5 94 49.8 118 57.5 113 60.5 120 23.2 96 31.0 105 15.7 73 23.2 101 49.6 90 19.3 91 30.5 108 39.0 102 19.0 94 9.25 90 18.6 91 9.82 75
Modified CLG [34]96.5 34.8 104 61.1 103 35.3 102 33.3 107 46.5 101 41.7 110 36.8 110 63.0 109 45.1 113 22.1 104 55.4 109 18.7 104 23.9 100 31.2 106 21.7 98 15.8 68 51.5 97 14.8 70 9.01 46 24.6 57 11.1 57 17.6 113 25.7 115 29.6 112
2D-CLG [1]98.0 44.0 111 63.3 105 36.1 103 44.3 114 52.3 111 55.1 118 49.1 119 75.4 117 50.5 117 64.3 123 76.4 122 67.8 123 24.8 104 29.7 95 27.4 110 20.5 89 53.6 102 22.4 96 2.52 3 13.0 2 3.50 14 22.8 118 27.9 118 36.9 117
Filter Flow [19]98.0 33.3 101 51.7 90 20.1 90 25.0 99 47.2 104 27.7 98 27.7 101 50.0 99 37.9 102 31.7 111 54.1 108 29.9 110 25.8 107 31.2 106 28.3 112 26.4 107 52.9 101 24.7 104 42.3 121 61.5 125 13.6 74 6.09 72 12.1 55 6.88 56
SPSA-learn [13]98.3 35.8 105 71.2 115 43.1 109 28.4 103 47.0 102 32.8 104 31.4 106 57.7 104 42.2 107 22.2 105 51.0 106 22.9 107 23.9 100 29.4 91 24.6 105 24.8 104 56.5 106 25.1 106 10.7 53 25.1 59 3.72 17 21.7 116 24.9 114 35.5 116
BlockOverlap [61]99.2 41.4 107 54.1 97 36.2 104 27.3 102 41.4 94 32.6 103 26.2 100 46.5 95 31.8 98 20.0 101 36.6 84 18.1 103 22.4 89 26.8 63 25.7 108 24.5 103 45.6 81 21.1 95 39.3 120 47.0 117 43.5 119 13.8 106 16.0 81 28.7 111
IAOF [50]100.1 33.8 102 58.3 99 40.6 105 33.0 106 44.5 98 39.5 108 30.6 105 58.7 106 33.8 100 34.1 114 52.4 107 40.8 114 23.1 95 29.9 97 19.7 90 22.5 97 53.7 103 16.8 82 22.3 92 34.0 86 10.0 52 19.5 114 23.9 111 37.1 119
GroupFlow [9]101.2 42.7 108 67.1 110 53.4 112 44.8 116 63.8 120 50.2 115 36.7 109 69.4 114 43.9 110 17.2 100 46.0 101 16.7 100 27.8 113 34.9 112 21.2 97 36.7 119 67.0 115 43.6 120 6.40 24 21.7 39 7.17 42 16.6 110 25.9 116 25.0 108
GraphCuts [14]101.4 34.5 103 59.0 100 32.1 101 26.2 100 51.1 109 26.4 97 28.1 102 51.7 100 40.4 106 13.0 98 47.5 102 7.98 96 23.7 99 30.0 99 24.3 103 33.4 116 45.2 79 25.7 108 31.2 110 37.7 99 36.8 111 10.7 97 19.7 98 17.7 100
Black & Anandan [4]101.6 38.5 106 69.5 113 53.4 112 28.5 104 49.6 107 32.0 102 33.4 107 60.5 107 40.2 105 22.6 106 55.9 110 22.8 106 24.5 103 32.2 110 19.6 88 21.7 94 58.9 110 22.7 97 22.6 93 37.6 98 5.27 33 16.7 111 22.6 108 25.4 109
2bit-BM-tele [98]104.4 55.1 118 64.1 108 69.1 119 21.4 95 38.7 77 25.3 96 23.3 97 46.7 96 21.2 92 26.2 107 38.7 91 25.2 108 24.9 105 29.9 97 27.7 111 31.3 114 52.6 100 34.6 114 43.3 122 51.7 122 54.5 123 10.3 96 20.1 99 16.8 99
Nguyen [33]105.8 43.9 110 66.0 109 42.8 108 54.0 121 49.4 106 70.1 123 42.9 114 67.4 112 47.3 115 55.4 121 65.7 117 64.4 121 27.0 111 31.8 109 31.0 115 22.8 98 54.8 104 27.3 109 13.1 64 25.2 60 6.08 38 22.2 117 26.9 117 38.9 120
SILK [79]108.0 49.5 117 69.2 112 69.3 121 39.9 112 60.6 115 47.0 113 40.4 113 70.7 115 45.6 114 32.0 112 56.5 111 31.2 112 31.4 116 36.9 117 33.3 117 31.1 113 63.2 113 32.3 113 10.3 50 23.0 45 17.3 83 25.0 120 31.9 120 36.9 117
UnFlow [129]108.2 70.9 123 78.9 118 58.5 115 51.3 120 67.4 123 56.9 119 54.4 121 83.6 124 52.8 120 33.4 113 60.2 114 30.1 111 36.7 123 38.4 120 46.2 124 38.2 120 69.6 119 42.8 119 26.2 102 40.3 104 1.60 5 7.20 80 18.3 88 8.75 71
Periodicity [78]109.8 48.9 116 63.9 107 41.0 106 34.5 108 60.0 114 37.5 107 55.4 122 67.4 112 56.6 122 20.4 102 56.9 112 17.5 101 53.2 126 66.7 127 46.5 125 48.3 124 76.0 125 46.4 122 9.14 47 34.4 88 9.98 51 28.3 123 48.2 126 40.6 122
Horn & Schunck [3]110.2 43.3 109 80.7 121 58.6 116 32.5 105 59.7 113 35.1 105 40.2 112 76.3 120 44.7 112 31.5 110 64.8 115 32.6 113 29.3 115 36.4 116 27.0 109 27.5 111 68.7 117 29.7 111 27.0 104 43.3 110 7.32 43 25.9 121 36.5 122 34.6 115
Heeger++ [104]112.8 61.9 121 80.2 120 47.4 111 44.8 116 77.8 126 40.9 109 68.0 127 84.7 125 62.1 125 43.6 117 69.1 118 41.9 115 32.6 118 37.9 118 32.0 116 51.1 125 78.4 126 54.2 125 13.3 66 43.4 111 10.4 54 15.0 107 21.4 106 19.6 104
SLK [47]113.5 44.7 112 78.9 118 59.1 117 58.2 123 71.2 124 70.9 124 47.5 117 83.5 123 50.6 118 65.0 124 69.5 119 73.4 124 34.7 120 38.9 122 42.9 123 34.8 117 70.9 122 39.4 117 12.1 59 29.8 73 11.5 58 34.4 124 40.1 123 48.8 124
TI-DOFE [24]115.6 73.1 124 84.6 125 89.6 127 61.2 125 64.7 122 74.8 126 58.6 123 88.7 127 58.0 123 70.9 125 81.6 124 76.1 125 31.7 117 38.0 119 35.4 118 29.7 112 68.7 117 36.3 115 17.1 79 32.0 79 8.67 48 35.5 125 42.8 124 49.8 125
FFV1MT [106]115.8 59.9 120 77.8 117 53.6 114 37.7 111 72.5 125 37.0 106 63.6 125 82.1 122 62.4 126 42.8 116 73.9 121 42.6 116 41.9 125 45.8 125 52.3 126 52.5 126 81.8 127 56.1 126 20.2 88 42.4 108 18.3 91 15.0 107 21.4 106 19.6 104
FOLKI [16]118.0 48.0 115 71.5 116 68.8 118 48.6 119 63.2 117 59.5 120 43.0 115 75.6 118 44.0 111 40.4 115 65.6 116 45.8 117 35.3 121 40.6 123 41.6 121 36.3 118 71.6 124 44.4 121 23.6 100 44.7 114 40.4 117 36.9 126 43.4 125 54.5 126
PGAM+LK [55]120.0 58.6 119 80.9 122 69.8 122 45.1 118 63.7 119 51.9 117 43.2 116 76.2 119 47.5 116 50.3 119 82.2 125 51.4 118 32.7 119 36.0 115 42.3 122 41.4 121 70.1 120 41.2 118 56.3 125 58.0 124 55.0 124 25.9 121 32.4 121 40.3 121
Adaptive flow [45]120.4 76.7 126 83.7 124 86.4 124 57.9 122 63.6 118 67.3 121 48.7 118 73.2 116 52.9 121 52.7 120 69.9 120 56.0 119 35.4 122 38.4 120 39.4 120 46.1 122 70.6 121 47.6 123 73.1 126 75.2 126 88.1 125 17.2 112 24.6 113 25.6 110
HCIC-L [99]122.2 76.5 125 86.4 126 73.3 123 70.1 126 62.5 116 85.3 127 63.5 124 66.1 111 79.5 127 83.3 127 91.8 127 86.5 127 39.0 124 42.8 124 38.7 119 46.4 123 66.6 114 52.3 124 89.6 127 85.9 127 94.0 126 19.8 115 24.2 112 29.6 112
Pyramid LK [2]123.7 68.1 122 83.5 123 86.8 125 59.4 124 64.5 121 73.1 125 52.8 120 76.3 120 61.4 124 60.2 122 79.0 123 65.9 122 53.8 127 61.8 126 64.5 127 59.4 127 71.1 123 63.0 127 43.9 123 49.4 121 39.5 115 50.2 127 60.2 127 70.8 127
AdaConv-v1 [126]128.0 100.0 128 99.9 128 100.0 128 100.0 128 100.0 128 100.0 128 99.9 128 99.9 128 99.9 128 100.0 128 100.0 128 99.9 128 99.9 128 99.9 128 99.9 128 99.9 128 99.8 128 100.0 128 99.7 128 97.0 128 99.9 128 99.9 128 99.9 128 99.9 128
SepConv-v1 [127]128.0 100.0 128 99.9 128 100.0 128 100.0 128 100.0 128 100.0 128 99.9 128 99.9 128 99.9 128 100.0 128 100.0 128 99.9 128 99.9 128 99.9 128 99.9 128 99.9 128 99.8 128 100.0 128 99.7 128 97.0 128 99.9 128 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.