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        
A75
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]16.0 2.17 4 5.35 5 1.92 4 1.56 5 6.39 19 1.67 10 1.51 7 3.67 8 1.61 8 1.20 36 4.36 17 1.02 37 2.30 4 3.18 3 1.73 7 2.16 39 6.35 8 2.39 65 2.54 25 4.18 13 2.17 31 0.81 9 1.50 17 0.67 4
NN-field [71]16.7 2.31 8 5.94 13 1.98 5 1.83 22 7.21 37 1.97 29 1.54 9 3.55 7 1.68 12 0.96 10 3.04 4 0.75 6 2.30 4 3.25 4 1.69 5 1.72 18 3.81 1 1.65 7 3.12 57 4.76 54 2.60 55 0.79 6 1.58 26 0.59 2
TC/T-Flow [76]20.0 2.06 1 7.42 30 1.55 1 1.61 11 6.77 28 1.52 3 1.46 5 4.33 18 1.56 5 0.88 3 6.96 45 0.69 2 2.91 20 4.35 26 1.88 11 1.47 5 6.12 7 1.61 6 2.22 8 3.98 4 4.00 101 1.06 41 1.99 46 1.19 53
ComponentFusion [96]20.2 2.12 3 6.18 16 1.80 3 1.67 14 4.82 4 1.91 24 1.34 2 4.01 11 1.47 3 0.88 3 4.74 19 0.72 4 2.99 24 4.35 26 1.96 14 2.03 31 11.2 76 1.96 27 2.90 47 4.62 48 1.90 18 0.93 26 1.56 25 0.89 16
ProFlow_ROB [147]20.4 2.44 11 7.85 39 2.01 6 1.54 3 7.07 33 1.55 4 1.53 8 7.16 55 1.43 2 0.81 1 5.26 26 0.67 1 3.48 43 4.95 48 1.69 5 1.46 4 8.54 34 1.53 4 1.92 3 4.48 33 2.01 22 0.93 26 2.14 52 0.94 26
ALD-Flow [66]23.3 2.26 6 5.81 10 2.07 7 1.56 5 5.71 13 1.66 8 1.46 5 4.64 24 1.66 11 0.95 9 7.15 49 0.80 12 3.18 34 4.57 37 1.76 8 1.51 6 7.63 20 1.60 5 2.61 28 4.27 18 3.90 97 1.04 38 2.22 59 1.18 50
WLIF-Flow [93]23.4 2.63 20 5.51 6 2.41 18 2.14 44 7.08 34 2.32 43 1.63 12 4.19 14 1.84 19 1.05 16 4.21 14 0.88 24 2.92 22 4.37 28 2.30 33 2.02 30 7.21 16 1.89 23 2.73 36 4.27 18 2.78 64 0.84 11 1.37 6 0.83 11
nLayers [57]24.5 2.33 9 5.14 3 2.17 10 2.75 92 7.22 39 3.07 94 1.69 21 4.04 12 2.21 64 0.88 3 2.83 2 0.70 3 2.08 2 3.25 4 1.30 1 1.88 21 6.06 6 1.85 14 2.89 46 4.59 43 2.28 36 0.92 23 1.48 15 0.94 26
OFLAF [77]25.0 2.77 36 5.70 8 2.49 20 1.76 17 5.35 9 1.84 21 1.54 9 2.69 2 1.72 14 1.30 41 3.55 10 1.12 51 2.30 4 3.62 7 1.64 4 2.23 45 5.93 3 2.06 37 2.81 40 4.29 21 2.99 70 1.15 51 1.69 31 1.18 50
MDP-Flow2 [68]25.5 3.08 41 6.23 17 2.73 46 1.55 4 4.80 3 1.64 7 1.63 12 3.27 4 1.61 8 1.37 52 5.15 24 1.15 56 2.91 20 4.19 20 2.20 21 2.24 47 6.43 9 2.17 47 2.62 29 4.35 25 1.88 17 1.08 43 1.66 30 0.95 30
RNLOD-Flow [121]26.3 2.19 5 4.96 2 2.19 11 1.79 19 7.20 36 1.76 16 1.42 3 4.31 16 1.56 5 0.97 11 3.42 9 0.84 17 2.75 15 4.16 17 2.01 16 1.86 20 7.24 17 1.95 26 4.02 95 6.41 112 4.48 116 0.90 18 1.51 18 0.85 12
OAR-Flow [125]27.3 2.55 14 7.57 32 2.36 15 1.81 20 7.94 45 1.94 28 1.72 28 8.40 62 1.95 27 0.94 8 5.90 36 0.79 9 3.49 45 4.83 46 1.84 10 1.16 2 7.84 25 1.22 2 1.93 4 3.63 1 2.26 34 1.10 46 2.16 55 1.34 62
AGIF+OF [85]28.5 2.60 15 6.24 18 2.45 19 2.49 72 9.30 67 2.63 71 1.68 19 4.79 30 2.08 48 1.05 16 4.34 16 0.81 13 2.77 17 4.05 14 2.10 19 1.92 24 7.48 19 1.74 12 2.69 33 4.48 33 2.69 61 0.87 12 1.40 7 0.95 30
Layers++ [37]28.9 2.70 31 6.40 20 2.83 53 2.33 61 6.62 25 2.54 68 1.65 16 3.24 3 2.02 37 0.92 6 2.48 1 0.75 6 2.12 3 3.11 1 1.50 3 2.06 34 8.25 28 1.94 25 3.59 80 5.41 81 3.21 73 0.89 15 1.32 4 0.90 19
LME [70]29.9 2.90 38 5.83 11 2.30 14 1.60 8 4.45 2 1.74 15 1.71 25 4.14 13 2.00 33 1.35 47 6.61 40 1.13 52 3.07 29 4.32 25 2.57 46 2.08 36 8.09 27 2.00 33 2.78 39 4.53 39 2.29 37 1.06 41 1.74 33 0.96 34
HAST [109]31.2 2.09 2 4.28 1 1.69 2 1.60 8 5.31 8 1.55 4 1.28 1 2.09 1 1.40 1 0.92 6 3.39 7 0.78 8 2.01 1 3.14 2 1.48 2 2.40 60 8.62 36 2.41 67 4.08 97 7.33 133 7.69 135 1.23 58 1.59 28 1.73 80
TC-Flow [46]31.5 2.45 12 6.60 22 2.39 17 1.25 1 5.24 6 1.32 1 1.45 4 4.40 21 1.50 4 1.15 33 8.13 56 1.04 39 3.22 35 4.77 44 2.06 18 1.94 25 8.65 37 2.09 41 2.33 19 4.51 36 3.82 95 1.24 59 2.18 57 1.50 75
PH-Flow [101]32.8 2.62 17 7.58 34 2.53 26 2.13 41 8.78 54 2.37 47 1.70 22 4.39 20 2.06 44 1.08 19 7.06 48 0.85 18 2.72 12 3.91 11 2.04 17 2.06 34 8.33 29 1.96 27 3.48 75 4.62 48 4.03 102 0.90 18 1.41 10 0.88 14
NNF-EAC [103]33.1 3.07 40 6.73 24 2.70 45 1.62 13 5.30 7 1.72 13 1.71 25 3.89 10 1.79 16 1.36 50 6.36 39 1.15 56 3.02 25 4.37 28 2.28 31 2.44 63 6.90 12 2.28 59 2.90 47 4.51 36 2.19 32 1.12 48 1.83 39 0.98 36
Classic+CPF [83]33.2 2.65 26 7.22 29 2.53 26 2.37 63 9.14 64 2.51 64 1.67 18 5.05 34 2.03 39 1.01 12 5.38 28 0.79 9 2.90 19 4.17 18 2.33 35 1.88 21 8.44 31 1.70 9 3.19 60 4.60 45 3.72 88 0.92 23 1.40 7 0.95 30
Sparse-NonSparse [56]33.3 2.62 17 7.58 34 2.60 35 2.18 50 8.74 52 2.46 60 1.68 19 4.86 31 2.00 33 1.04 14 7.97 54 0.81 13 3.13 31 4.45 32 2.42 40 1.98 28 8.53 33 1.87 18 3.13 58 4.32 24 3.51 81 0.88 13 1.41 10 0.91 20
FC-2Layers-FF [74]33.6 2.63 20 5.87 12 2.68 43 2.19 54 8.07 47 2.39 51 1.65 16 3.42 5 2.05 41 1.10 25 3.11 5 0.88 24 2.57 7 3.49 6 2.30 33 2.26 50 7.68 22 2.17 47 3.70 83 5.18 75 3.73 89 0.90 18 1.41 10 0.93 24
IROF++ [58]33.9 2.66 27 6.82 25 2.58 33 2.17 48 9.07 61 2.41 53 1.75 34 5.03 33 2.06 44 1.11 27 7.65 51 0.90 30 2.92 22 4.18 19 2.25 28 2.13 37 9.85 55 1.98 30 2.53 24 4.53 39 1.43 6 0.97 35 1.58 26 0.94 26
FESL [72]36.5 2.63 20 5.15 4 2.77 50 2.62 85 9.27 66 2.73 78 1.72 28 4.77 28 2.07 47 1.15 33 3.36 6 0.98 35 2.75 15 3.93 12 2.20 21 1.95 26 7.12 14 1.98 30 3.40 70 5.71 93 2.89 66 0.88 13 1.55 22 0.87 13
JOF [141]36.6 2.52 13 6.01 15 2.37 16 2.40 67 9.13 63 2.66 73 1.63 12 3.82 9 2.17 61 1.02 13 5.46 31 0.79 9 2.69 11 3.93 12 2.24 25 2.16 39 8.05 26 2.20 49 3.97 93 5.74 94 5.45 122 0.82 10 1.33 5 0.82 10
COFM [59]36.6 2.28 7 7.19 27 2.08 9 1.78 18 6.57 22 1.93 26 1.56 11 5.33 39 2.19 62 0.86 2 4.90 20 0.73 5 3.76 55 4.80 45 3.87 95 2.03 31 7.67 21 1.72 10 2.76 37 4.21 14 4.04 104 1.62 86 1.87 41 2.04 92
Efficient-NL [60]37.3 2.38 10 5.67 7 2.07 7 2.45 71 8.54 49 2.55 69 1.64 15 4.63 23 1.95 27 1.04 14 5.76 34 0.81 13 2.74 13 4.14 16 1.94 13 2.87 77 8.58 35 2.23 52 3.30 65 5.12 72 3.05 71 1.15 51 1.83 39 1.19 53
PMMST [114]38.3 3.55 64 6.48 21 3.33 74 2.18 50 6.49 21 2.47 63 1.93 46 4.28 15 2.09 49 1.60 71 2.84 3 1.43 79 2.60 8 3.72 9 1.91 12 2.22 44 6.05 5 2.13 45 2.70 35 4.48 33 2.08 26 1.18 57 1.91 43 1.09 46
LSM [39]38.9 2.60 15 7.70 38 2.61 37 2.19 54 8.77 53 2.44 58 1.70 22 4.78 29 2.06 44 1.08 19 8.13 56 0.86 20 3.05 27 4.30 23 2.47 41 2.18 42 8.66 40 2.08 39 3.56 78 4.68 53 3.74 90 0.90 18 1.43 13 0.93 24
Classic+NL [31]40.1 2.63 20 7.57 32 2.64 40 2.18 50 9.04 59 2.41 53 1.71 25 4.70 25 2.09 49 1.08 19 7.69 52 0.88 24 3.04 26 4.31 24 2.41 39 2.27 51 8.65 37 2.09 41 3.79 87 5.12 72 3.81 93 0.89 15 1.44 14 0.89 16
FMOF [94]40.6 2.71 33 6.71 23 2.62 39 2.64 86 9.19 65 2.77 79 1.73 31 4.37 19 2.22 65 1.06 18 5.18 25 0.81 13 2.89 18 4.25 22 2.40 38 2.31 53 7.76 24 1.96 27 3.35 68 4.88 66 3.81 93 0.93 26 1.55 22 0.92 22
Ramp [62]40.7 2.64 25 7.64 37 2.56 30 2.20 57 8.90 57 2.46 60 1.73 31 4.74 27 2.09 49 1.11 27 6.81 44 0.88 24 3.07 29 4.46 33 2.38 37 2.28 52 8.52 32 2.15 46 3.40 70 4.25 15 4.16 111 0.94 29 1.53 20 0.97 35
ProbFlowFields [128]41.7 3.31 51 13.8 88 2.85 55 2.03 33 6.67 27 2.23 38 1.89 44 6.06 45 2.29 69 1.23 38 5.14 23 0.95 33 3.70 51 5.08 50 2.29 32 1.60 10 7.44 18 1.85 14 2.43 23 4.30 22 2.45 47 1.29 63 2.32 61 1.38 66
2DHMM-SAS [92]43.4 2.62 17 7.61 36 2.53 26 2.17 48 10.1 77 2.38 49 1.83 39 6.16 48 2.10 53 1.09 23 8.07 55 0.86 20 3.06 28 4.42 31 2.37 36 2.16 39 9.30 46 2.01 34 3.50 76 4.64 50 4.14 110 0.96 33 1.64 29 0.99 39
Adaptive [20]43.9 2.63 20 8.08 43 2.23 12 2.18 50 8.66 51 2.27 41 2.04 55 9.57 69 2.09 49 1.12 30 10.6 75 0.87 22 4.47 100 5.30 67 4.44 111 1.55 7 8.65 37 1.43 3 3.32 66 5.51 83 2.14 30 0.77 5 1.52 19 0.75 8
SVFilterOh [111]44.6 3.48 60 5.72 9 3.45 79 2.38 65 6.06 16 2.41 53 1.93 46 3.46 6 2.05 41 1.53 67 3.71 11 1.27 67 2.65 10 4.10 15 2.00 15 2.47 65 7.08 13 2.28 59 4.60 108 7.10 126 5.92 127 0.73 4 1.16 2 0.70 7
PMF [73]45.2 3.22 49 6.34 19 2.60 35 1.95 29 6.66 26 1.92 25 1.85 40 4.58 22 1.83 18 1.62 72 4.27 15 1.37 73 2.61 9 3.74 10 1.83 9 3.18 86 9.94 58 3.34 92 5.29 124 8.26 139 5.58 126 0.68 3 1.21 3 0.67 4
TV-L1-MCT [64]45.9 2.68 29 6.83 26 2.61 37 2.68 87 10.3 80 2.80 83 1.78 36 5.24 36 2.24 66 1.13 31 5.44 30 0.87 22 3.39 41 4.69 42 2.96 64 2.45 64 9.14 44 2.31 63 2.64 30 4.37 27 2.04 24 1.08 43 1.74 33 1.36 64
S2D-Matching [84]46.0 2.69 30 7.86 41 2.68 43 2.19 54 9.11 62 2.44 58 1.77 35 6.11 47 2.04 40 1.13 31 5.86 35 0.91 31 3.17 33 4.41 30 2.50 43 2.41 61 9.09 43 2.25 56 4.00 94 5.16 74 4.07 105 0.91 22 1.40 7 0.95 30
SimpleFlow [49]47.8 2.74 34 8.28 47 2.73 46 2.50 73 9.73 75 2.83 88 1.89 44 6.81 53 2.35 72 1.11 27 10.4 72 0.89 29 3.27 36 4.47 34 2.63 48 3.03 80 8.91 42 2.39 65 3.10 56 4.25 15 2.76 63 0.89 15 1.49 16 0.89 16
Occlusion-TV-L1 [63]48.2 3.15 43 8.42 49 2.50 21 2.03 33 7.42 41 2.14 35 2.24 66 9.79 71 2.16 59 1.35 47 9.59 65 1.11 47 4.10 78 5.77 98 3.22 75 1.68 17 9.21 45 2.08 39 2.69 33 4.59 43 1.70 13 1.05 39 2.36 64 0.98 36
IROF-TV [53]48.4 2.89 37 8.67 52 2.81 52 2.25 59 9.54 73 2.51 64 1.79 38 5.50 41 2.15 58 1.53 67 11.5 83 1.27 67 3.33 37 4.62 38 2.85 56 2.78 72 13.5 98 2.57 70 2.15 6 4.14 11 1.37 5 0.94 29 1.55 22 0.94 26
MDP-Flow [26]49.0 3.14 42 9.81 60 2.83 53 2.06 36 6.10 17 2.43 56 1.87 42 6.10 46 2.10 53 1.44 57 8.90 61 1.15 56 3.37 39 4.62 38 2.54 45 2.35 56 10.4 64 2.23 52 2.88 44 4.83 59 1.94 20 1.27 62 2.62 71 1.09 46
Correlation Flow [75]49.6 3.18 45 7.85 39 2.85 55 1.74 15 5.77 14 1.69 12 1.94 48 5.25 37 1.70 13 1.47 61 5.67 33 1.26 66 3.66 48 5.20 61 2.53 44 3.06 82 9.57 50 3.10 86 3.42 73 4.94 68 4.03 102 1.17 55 1.80 36 1.16 48
3DFlow [135]50.2 3.21 47 7.43 31 2.64 40 1.92 27 7.04 30 1.85 22 2.03 54 4.31 16 1.84 19 1.65 78 3.41 8 1.33 69 3.14 32 4.51 35 2.24 25 3.66 94 11.8 82 3.87 105 4.19 100 4.93 67 5.46 123 1.05 39 1.54 21 1.02 42
AggregFlow [97]52.0 3.29 50 8.49 50 3.14 66 2.70 89 12.2 98 2.68 75 2.32 70 9.03 64 2.88 88 1.44 57 4.19 13 1.25 65 3.48 43 5.09 51 2.19 20 1.55 7 5.36 2 1.68 8 2.56 27 4.78 56 1.77 15 1.54 82 2.15 53 2.16 98
IIOF-NLDP [131]52.5 3.19 46 10.4 64 2.50 21 2.43 70 9.32 69 2.32 43 1.98 51 5.55 42 1.81 17 1.48 63 6.12 38 1.23 63 3.75 53 5.55 86 2.25 28 3.03 80 9.30 46 3.02 82 2.66 32 4.83 59 2.60 55 1.24 59 1.97 45 1.17 49
Aniso-Texture [82]53.9 2.75 35 6.00 14 3.09 63 2.13 41 5.64 11 2.52 67 1.78 36 6.80 52 2.20 63 1.08 19 4.01 12 0.92 32 4.08 75 5.44 77 3.26 79 2.31 53 11.8 82 2.26 57 5.87 132 8.09 138 4.24 112 0.80 8 1.70 32 0.67 4
OFH [38]54.5 3.60 72 10.3 63 3.80 90 1.58 7 7.05 31 1.66 8 1.70 22 9.23 65 1.58 7 1.19 35 10.1 68 1.08 43 3.98 61 5.22 63 3.57 84 2.80 73 12.6 89 3.12 87 2.30 15 4.60 45 2.35 39 1.41 76 2.90 84 1.75 81
CostFilter [40]55.7 3.59 69 8.35 48 3.26 70 2.12 39 6.60 23 2.16 36 2.00 53 5.56 43 2.01 35 2.04 91 7.05 47 1.89 92 2.74 13 3.70 8 2.27 30 3.29 87 10.3 63 3.33 91 5.22 122 9.79 144 6.16 129 0.38 1 1.08 1 0.35 1
Classic++ [32]55.8 2.66 27 8.18 46 2.65 42 2.13 41 7.96 46 2.43 56 1.85 40 9.39 68 2.10 53 1.09 23 10.4 72 0.88 24 3.97 59 5.60 91 2.89 59 2.36 57 13.6 101 2.10 43 4.03 96 5.20 76 4.33 113 0.99 37 2.05 48 0.92 22
DeepFlow2 [108]56.2 3.44 56 12.6 78 3.36 77 1.98 30 8.60 50 2.10 33 2.38 73 11.1 76 2.60 79 1.34 44 14.6 94 1.11 47 3.53 46 5.09 51 2.23 24 1.66 16 9.90 57 1.77 13 2.76 37 4.06 8 3.40 78 1.72 91 3.21 95 2.13 96
S2F-IF [123]56.7 3.54 63 19.2 115 2.58 33 2.41 68 10.4 81 2.59 70 2.57 78 10.6 75 2.59 77 1.29 40 10.5 74 0.98 35 4.07 73 5.38 72 2.90 60 1.65 15 9.94 58 1.85 14 2.27 10 4.26 17 2.35 39 1.33 66 2.49 68 1.32 59
FlowFields+ [130]58.2 3.58 67 19.0 111 2.56 30 2.57 78 10.8 86 2.78 80 2.72 84 11.9 82 2.78 86 1.32 42 10.9 78 1.03 38 3.97 59 5.34 69 2.74 49 1.64 14 9.79 54 1.87 18 2.26 9 4.30 22 2.35 39 1.35 68 2.62 71 1.34 62
CPM-Flow [116]59.0 3.47 58 19.0 111 2.52 23 2.59 80 11.0 91 2.82 85 2.56 76 11.3 78 2.75 82 1.34 44 15.7 99 1.06 40 4.02 66 5.42 74 2.78 51 1.59 9 9.39 48 1.87 18 2.30 15 4.17 12 2.36 42 1.35 68 2.69 77 1.39 68
RFlow [90]59.5 3.62 74 9.91 61 3.53 83 1.83 22 5.50 10 1.93 26 2.14 61 9.57 69 1.86 22 1.32 42 6.75 41 1.14 54 3.98 61 5.35 70 3.24 76 2.39 59 11.7 81 2.24 54 3.45 74 4.60 45 3.63 83 1.64 88 2.90 84 1.91 88
PGM-C [120]59.7 3.48 60 19.0 111 2.52 23 2.59 80 10.8 86 2.82 85 2.59 79 11.8 81 2.75 82 1.34 44 16.5 104 1.06 40 4.03 68 5.46 79 2.78 51 1.60 10 9.63 52 1.88 21 2.28 11 3.98 4 2.36 42 1.37 71 2.69 77 1.45 71
EpicFlow [102]61.4 3.47 58 18.9 109 2.52 23 2.59 80 10.9 89 2.82 85 2.64 81 14.2 92 2.75 82 1.35 47 15.5 97 1.06 40 4.04 70 5.48 80 2.88 58 1.62 13 9.70 53 1.91 24 2.28 11 4.08 10 2.36 42 1.39 75 2.71 79 1.51 76
FlowFields [110]61.8 3.56 66 19.0 111 2.54 29 2.57 78 10.6 84 2.79 81 2.72 84 11.7 80 2.76 85 1.42 54 10.9 78 1.13 52 4.08 75 5.43 76 2.92 62 1.60 10 10.5 66 1.86 17 2.28 11 4.35 25 2.46 49 1.38 74 2.62 71 1.36 64
WRT [151]62.1 3.48 60 8.79 54 2.57 32 3.21 99 9.36 71 3.18 95 2.76 87 7.40 57 2.30 70 1.64 77 4.61 18 1.23 63 3.41 42 4.56 36 2.48 42 4.89 119 11.0 73 3.31 89 3.03 53 4.82 58 3.25 75 1.10 46 1.75 35 0.99 39
MLDP_OF [89]62.2 4.16 92 10.4 64 4.04 92 2.04 35 6.61 24 2.04 31 2.36 72 6.60 51 2.05 41 1.43 55 5.65 32 1.18 60 3.75 53 4.86 47 2.96 64 2.96 78 8.71 41 3.47 94 4.20 101 5.51 83 7.24 132 1.16 54 1.87 41 1.21 55
DMF_ROB [140]63.9 3.70 76 15.2 97 3.34 75 2.21 58 9.03 58 2.40 52 2.84 90 13.7 91 2.65 81 1.41 53 16.9 107 1.10 45 3.92 57 5.17 59 3.14 72 2.01 29 10.6 67 2.11 44 2.37 21 3.72 3 2.62 58 1.49 81 2.74 80 1.68 79
TV-L1-improved [17]64.2 2.70 31 9.05 55 2.29 13 1.85 24 7.06 32 1.97 29 1.94 48 9.28 67 1.90 26 1.10 25 8.96 62 0.85 18 4.07 73 5.60 91 2.75 50 5.44 126 17.3 118 6.29 129 4.75 116 6.82 118 4.73 119 1.13 49 2.68 76 1.06 45
Steered-L1 [118]64.5 3.21 47 8.15 45 3.11 65 1.39 2 4.13 1 1.51 2 1.73 31 5.20 35 1.65 10 1.28 39 10.3 71 1.11 47 4.05 72 5.35 70 3.55 82 3.10 83 12.6 89 2.59 71 6.15 137 6.90 120 13.1 144 1.73 92 3.04 92 2.39 102
BriefMatch [124]64.6 3.03 39 7.96 42 2.73 46 1.75 16 6.88 29 1.73 14 1.72 28 4.70 25 1.73 15 1.51 66 5.38 28 1.39 77 4.01 64 5.27 65 3.72 90 5.57 127 15.9 111 6.02 128 4.65 109 6.85 119 8.98 139 0.95 31 2.30 60 1.77 82
CombBMOF [113]65.3 3.55 64 11.6 73 2.79 51 2.52 74 7.21 37 2.51 64 1.88 43 5.63 44 1.84 19 1.67 81 11.2 81 1.51 84 3.68 49 4.62 38 3.07 70 4.08 101 11.4 79 4.79 118 4.72 113 6.58 115 3.82 95 0.92 23 1.82 38 0.88 14
PWC-Net_ROB [148]65.3 4.74 104 14.6 92 3.51 82 2.88 95 8.88 56 2.92 92 2.75 86 9.98 73 3.27 96 1.65 78 4.91 21 1.35 70 4.10 78 5.12 56 2.91 61 2.66 70 10.2 62 2.67 72 1.69 2 4.65 52 1.13 2 1.26 61 2.06 49 1.29 58
Sparse Occlusion [54]66.2 3.36 52 8.08 43 2.90 57 2.61 84 7.68 42 3.01 93 2.10 57 6.40 50 2.13 56 1.45 59 6.80 42 1.14 54 4.01 64 5.31 68 2.81 54 2.55 67 10.4 64 2.21 51 6.70 140 8.26 139 4.34 114 1.15 51 2.08 50 0.99 39
DeepFlow [86]67.3 3.94 84 12.7 79 4.14 96 2.12 39 9.06 60 2.28 42 2.84 90 12.5 86 3.16 93 1.68 82 15.6 98 1.44 81 3.58 47 5.10 53 2.20 21 1.78 19 11.1 74 1.88 21 2.65 31 4.07 9 3.40 78 2.08 109 3.59 110 3.09 113
EPPM w/o HM [88]68.0 4.03 89 13.7 87 3.25 69 1.91 26 7.71 43 1.83 19 2.14 61 7.85 60 1.96 29 1.80 84 10.2 69 1.63 88 3.72 52 4.62 38 3.24 76 3.93 99 13.2 96 3.79 103 4.35 105 5.68 92 7.45 133 0.97 35 1.93 44 0.98 36
Complementary OF [21]70.7 4.47 100 12.4 77 4.63 105 1.60 8 6.16 18 1.67 10 2.10 57 6.85 54 2.16 59 2.27 96 9.76 66 2.19 101 4.00 63 5.10 53 3.71 88 3.96 100 12.9 93 3.32 90 2.83 41 4.46 32 3.08 72 2.04 107 3.33 100 2.86 107
FF++_ROB [146]71.0 3.75 78 20.6 117 2.92 58 2.60 83 10.6 84 2.79 81 2.97 97 13.2 90 3.18 94 1.62 72 11.2 81 1.38 76 4.12 80 5.52 84 2.99 67 2.24 47 9.58 51 2.30 62 2.31 17 4.40 30 2.39 45 1.36 70 2.55 69 1.44 70
HBM-GC [105]71.1 5.52 108 7.21 28 5.03 112 2.96 96 7.15 35 3.23 97 2.79 89 4.90 32 2.88 88 3.12 111 4.92 22 2.97 114 3.37 39 4.23 21 3.46 80 3.80 98 6.63 10 3.52 95 5.86 131 7.23 131 4.53 117 0.64 2 2.02 47 0.64 3
TF+OM [100]72.5 3.58 67 9.07 56 2.75 49 2.07 37 6.43 20 2.37 47 1.99 52 7.56 59 2.78 86 2.07 92 7.02 46 2.07 99 4.19 85 5.12 56 4.32 106 3.15 85 10.1 61 3.00 81 4.10 98 6.00 100 3.92 98 1.54 82 2.98 89 1.94 89
Rannacher [23]72.7 3.60 72 11.3 70 3.27 71 2.41 68 9.53 72 2.63 71 2.60 80 11.9 82 2.58 76 1.36 50 12.1 85 1.09 44 4.22 86 5.90 105 3.14 72 3.63 93 16.1 114 2.75 76 3.72 84 5.24 78 3.70 87 0.96 33 2.16 55 0.91 20
Aniso. Huber-L1 [22]73.6 3.17 44 9.57 58 3.05 61 3.72 101 11.5 95 4.38 104 2.86 93 10.5 74 3.80 100 1.70 83 11.6 84 1.42 78 4.04 70 5.58 88 2.98 66 2.34 55 9.88 56 2.05 36 4.49 107 5.91 98 3.42 80 1.08 43 2.10 51 1.02 42
F-TV-L1 [15]74.2 5.69 110 13.3 82 6.62 119 2.71 90 12.0 97 2.86 90 2.76 87 12.6 87 2.43 74 2.41 101 16.3 103 2.02 98 4.17 83 5.27 65 3.74 91 2.41 61 10.8 71 2.49 69 3.04 54 4.84 64 2.26 34 0.79 6 1.81 37 0.76 9
TCOF [69]74.5 3.95 85 11.0 68 4.20 97 2.56 77 9.31 68 2.68 75 2.71 83 12.8 89 3.34 97 2.30 97 6.80 42 2.33 103 4.50 102 6.28 122 2.58 47 1.89 23 6.02 4 2.06 37 4.72 113 6.30 105 2.58 52 1.37 71 2.63 74 1.22 57
ComplOF-FED-GPU [35]74.5 4.09 91 12.8 80 4.09 95 1.61 11 9.86 76 1.62 6 2.12 59 8.39 61 1.87 24 1.85 86 12.4 88 1.70 91 3.95 58 5.25 64 3.25 78 3.54 91 15.1 109 3.60 99 3.93 90 4.83 59 4.60 118 1.55 84 2.93 88 1.80 83
NL-TV-NCC [25]75.0 3.89 82 8.49 50 3.34 75 2.52 74 8.44 48 2.38 49 2.25 67 5.49 40 1.99 31 1.87 87 7.61 50 1.53 85 4.36 93 5.91 106 2.78 51 4.12 105 13.0 94 3.58 98 3.85 88 5.74 94 3.79 92 1.63 87 2.81 82 1.46 72
ACK-Prior [27]75.7 4.28 94 9.53 57 3.85 91 1.87 25 5.68 12 1.83 19 1.97 50 5.25 37 1.96 29 1.98 90 5.26 26 1.65 89 4.08 75 5.12 56 3.79 94 4.53 115 13.0 94 3.61 100 5.63 126 6.40 110 8.50 137 1.92 103 2.90 84 2.64 104
ROF-ND [107]75.9 4.03 89 11.1 69 3.48 81 2.33 61 5.09 5 2.22 37 2.19 64 6.22 49 2.02 37 2.30 97 5.92 37 1.68 90 4.23 87 6.03 112 2.94 63 3.70 96 12.2 85 3.05 84 6.21 138 6.93 122 5.53 124 1.43 78 2.21 58 1.32 59
LDOF [28]80.5 3.72 77 14.9 95 3.59 86 2.38 65 14.0 109 2.46 60 2.69 82 14.4 93 2.55 75 1.48 63 33.9 128 1.10 45 4.24 89 5.59 90 3.75 92 2.04 33 16.4 116 1.99 32 2.83 41 4.83 59 2.43 46 2.28 117 4.02 122 3.38 116
CRTflow [80]80.5 3.65 75 14.0 90 3.10 64 2.16 46 7.88 44 2.23 38 2.25 67 11.2 77 1.99 31 1.56 69 12.7 89 1.35 70 4.02 66 5.53 85 3.01 68 6.86 133 19.6 129 8.64 135 3.29 63 5.53 86 3.23 74 2.05 108 3.95 121 2.71 105
DPOF [18]81.2 4.32 95 16.2 99 3.30 72 2.69 88 10.2 78 2.69 77 2.44 75 7.17 56 2.61 80 1.95 89 10.2 69 1.55 86 3.85 56 5.20 61 3.03 69 2.84 75 11.1 74 2.67 72 4.71 112 4.83 59 8.84 138 1.73 92 3.03 91 1.86 85
SRR-TVOF-NL [91]81.2 4.62 103 12.2 76 3.55 84 2.32 60 10.8 86 2.34 45 2.56 76 12.4 85 2.59 77 1.49 65 8.56 59 1.17 59 4.12 80 5.10 53 3.51 81 2.63 68 10.9 72 2.26 57 5.64 128 6.92 121 4.13 109 2.19 114 2.87 83 2.86 107
LocallyOriented [52]81.5 3.46 57 14.6 92 3.01 60 2.84 94 13.3 103 2.85 89 2.92 95 17.6 102 3.05 92 1.63 75 10.6 75 1.43 79 4.23 87 5.79 99 3.19 74 2.48 66 7.69 23 2.87 78 3.41 72 5.63 90 3.25 75 1.65 89 3.48 104 1.86 85
SIOF [67]82.2 4.00 87 8.74 53 3.46 80 2.00 32 13.6 104 2.13 34 3.02 100 15.7 98 3.38 98 2.55 105 13.5 93 2.50 106 4.27 90 5.70 94 3.70 87 3.55 92 11.5 80 4.01 106 3.17 59 4.64 50 2.12 29 1.85 101 3.29 99 2.15 97
Second-order prior [8]83.0 3.40 53 13.6 84 3.19 67 2.16 46 13.8 108 2.34 45 2.43 74 17.1 100 2.26 67 1.20 36 15.7 99 0.96 34 4.44 97 6.10 116 3.08 71 3.41 90 19.7 130 2.67 72 5.42 125 6.02 102 5.40 121 1.44 79 3.44 103 1.48 73
Brox et al. [5]83.2 4.01 88 14.7 94 4.49 103 2.75 92 11.5 95 3.21 96 2.33 71 12.2 84 2.34 71 1.46 60 19.9 111 1.19 61 4.62 107 5.71 95 4.89 119 2.13 37 13.3 97 2.28 59 2.87 43 4.78 56 1.55 10 2.30 119 3.68 114 3.31 114
Bartels [41]84.0 4.23 93 10.7 67 4.70 108 2.37 63 5.83 15 2.66 73 2.21 65 7.42 58 2.42 73 2.59 106 8.46 58 2.53 107 4.33 92 5.50 82 4.37 108 3.69 95 14.6 107 4.80 119 4.75 116 6.30 105 7.59 134 1.13 49 2.33 63 1.32 59
Dynamic MRF [7]84.5 4.55 102 13.6 84 5.02 110 1.81 20 8.86 55 1.82 18 2.13 60 12.6 87 1.87 24 1.62 72 13.2 91 1.45 82 4.61 105 5.80 102 4.32 106 4.14 106 21.3 132 4.42 111 3.22 62 4.41 31 5.01 120 2.11 110 3.92 120 3.51 117
CLG-TV [48]85.7 3.59 69 9.91 61 3.24 68 4.16 106 11.1 92 4.96 106 3.12 101 11.5 79 3.97 101 2.31 99 13.0 90 1.99 97 4.56 104 6.11 117 3.95 97 2.85 76 12.2 85 2.78 77 4.23 104 5.87 97 2.86 65 1.17 55 2.45 67 1.04 44
TriangleFlow [30]86.0 3.96 86 11.5 71 4.08 94 2.14 44 10.2 78 2.07 32 2.16 63 9.80 72 1.86 22 1.47 61 9.22 63 1.11 47 5.37 128 7.25 135 4.72 115 4.49 114 13.7 103 4.62 116 3.78 86 7.33 133 4.11 107 1.73 92 3.48 104 2.30 99
p-harmonic [29]86.6 4.47 100 14.4 91 4.52 104 2.71 90 9.33 70 2.89 91 3.40 102 15.0 95 3.02 91 1.93 88 24.1 118 1.59 87 4.15 82 5.18 60 3.66 86 3.37 89 16.0 112 3.54 96 3.90 89 5.36 80 2.71 62 1.29 63 2.41 66 1.38 66
Local-TV-L1 [65]87.0 4.95 105 13.2 81 5.40 113 4.37 108 14.6 111 5.04 108 4.59 109 17.8 104 5.96 110 2.42 102 16.9 107 2.25 102 3.68 49 5.03 49 2.82 55 2.25 49 10.6 67 2.24 54 2.55 26 4.37 27 2.91 69 2.73 126 4.10 124 7.77 132
FlowNetS+ft+v [112]87.1 3.42 55 13.4 83 3.39 78 2.54 76 11.2 93 2.80 83 2.94 96 18.6 108 4.76 103 1.43 55 27.4 121 1.20 62 4.67 110 6.35 124 3.71 88 1.96 27 12.3 87 2.01 34 4.10 98 6.00 100 4.11 107 1.76 95 3.49 107 2.37 101
CNN-flow-warp+ref [117]87.9 3.90 83 19.8 116 3.73 89 3.40 100 10.9 89 4.21 102 3.85 106 23.8 120 6.07 111 1.65 78 22.4 116 1.37 73 4.38 96 5.58 88 4.08 100 2.23 45 13.8 104 2.33 64 2.41 22 4.27 18 2.24 33 2.43 123 3.66 113 3.64 121
CBF [12]88.0 3.59 69 10.5 66 3.68 88 4.72 110 10.4 81 6.02 115 2.28 69 9.24 66 2.96 90 1.63 75 12.2 86 1.36 72 4.48 101 5.75 97 3.99 99 2.70 71 10.7 70 2.48 68 6.13 136 7.02 123 5.92 127 1.45 80 2.65 75 1.66 78
DF-Auto [115]88.0 3.88 81 17.6 103 2.93 59 5.44 116 14.7 112 6.44 116 4.54 108 16.8 99 9.38 116 2.22 95 15.0 96 1.95 94 4.32 91 6.00 110 3.88 96 1.44 3 7.14 15 1.73 11 4.20 101 6.78 117 1.70 13 2.42 122 4.04 123 3.34 115
SuperFlow [81]90.3 3.40 53 11.5 71 3.31 73 3.97 103 12.2 98 5.00 107 3.01 99 17.9 105 7.70 114 2.66 107 18.1 110 2.64 110 4.17 83 5.42 74 4.16 104 2.19 43 10.6 67 2.20 49 4.22 103 6.11 104 2.45 47 2.12 112 3.64 112 3.59 119
OFRF [134]91.5 4.35 96 9.74 59 4.34 98 5.12 114 13.1 102 5.68 113 3.74 105 14.7 94 5.10 104 2.91 109 11.0 80 2.84 111 3.34 38 4.74 43 2.24 25 3.34 88 9.46 49 3.23 88 3.67 82 5.54 87 5.56 125 3.05 130 3.88 118 9.53 137
StereoFlow [44]94.6 21.8 146 37.8 141 27.4 146 24.3 144 37.6 146 22.4 142 28.3 146 39.2 141 28.8 141 24.0 144 47.7 138 21.6 143 5.15 123 5.49 81 6.01 134 0.95 1 6.87 11 1.06 1 1.68 1 3.70 2 0.92 1 1.29 63 2.32 61 1.49 74
LiteFlowNet [143]94.8 6.43 113 24.0 123 4.45 101 3.74 102 10.4 81 3.78 99 4.32 107 15.2 97 3.78 99 2.36 100 8.74 60 1.97 96 4.88 117 6.06 113 4.38 109 4.23 109 14.0 105 3.68 101 3.56 78 5.59 89 1.98 21 1.70 90 2.79 81 1.83 84
TriFlow [95]95.5 4.44 97 13.8 88 3.62 87 3.16 98 9.65 74 3.81 100 2.89 94 19.6 111 6.36 112 2.48 104 7.88 53 2.33 103 4.37 95 5.44 77 4.28 105 2.98 79 8.42 30 3.07 85 11.7 145 7.70 136 21.5 145 1.76 95 2.98 89 1.94 89
Learning Flow [11]95.9 3.80 80 11.9 74 3.58 85 3.02 97 13.0 101 3.34 98 2.84 90 17.9 105 3.18 94 1.82 85 34.6 131 1.50 83 5.44 131 7.32 136 4.61 113 3.10 83 18.8 125 3.04 83 3.94 91 6.38 109 3.65 85 1.37 71 3.38 101 1.18 50
Fusion [6]96.1 3.76 79 16.9 100 4.07 93 1.99 31 7.37 40 2.26 40 2.07 56 8.51 63 2.28 68 1.59 70 24.8 119 1.37 73 5.00 121 6.36 125 4.98 124 4.70 118 16.2 115 5.01 122 6.00 135 7.50 135 4.38 115 2.97 129 3.74 117 3.55 118
ContinualFlow_ROB [153]97.3 7.09 119 26.0 127 5.87 116 6.48 120 13.6 104 7.04 120 7.43 119 21.2 115 9.67 117 3.12 111 10.8 77 2.57 109 5.41 130 6.38 126 4.71 114 6.48 132 15.7 110 7.64 133 2.20 7 4.02 7 1.33 4 1.42 77 2.40 65 1.64 77
Shiralkar [42]97.4 4.46 98 18.3 106 4.36 99 1.93 28 16.4 114 1.87 23 2.99 98 17.6 102 2.01 35 2.10 93 21.0 113 1.96 95 4.36 93 5.72 96 3.55 82 5.65 128 19.4 127 5.11 125 4.90 120 5.57 88 7.14 131 2.11 110 4.71 130 2.53 103
StereoOF-V1MT [119]100.0 4.46 98 18.0 105 4.46 102 2.09 38 18.6 120 1.79 17 3.70 104 20.6 114 2.13 56 2.18 94 25.0 120 1.91 93 5.52 133 6.98 133 4.82 117 5.01 124 25.8 137 4.73 117 3.21 61 5.27 79 3.64 84 2.32 120 4.64 128 2.83 106
EAI-Flow [152]101.4 7.60 121 21.3 118 6.38 118 4.08 104 15.8 113 4.21 102 5.17 111 18.4 107 5.60 108 3.05 110 15.7 99 2.94 113 4.45 98 5.81 103 3.60 85 4.33 110 12.6 89 4.10 107 5.63 126 6.30 105 3.58 82 1.33 66 2.55 69 1.43 69
SegOF [10]101.5 5.62 109 17.1 101 3.08 62 8.33 126 20.9 122 10.1 129 7.44 120 21.7 116 13.3 125 5.42 127 21.0 113 4.47 122 4.81 116 5.51 83 5.74 133 4.97 122 17.1 117 4.83 120 2.12 5 4.38 29 1.46 7 2.17 113 3.23 96 3.74 123
WOLF_ROB [149]103.0 5.28 107 22.2 120 4.65 107 4.15 105 21.9 126 3.81 100 6.02 115 23.6 119 5.35 107 2.47 103 16.5 104 2.33 103 4.50 102 5.65 93 4.12 102 4.15 107 14.8 108 3.83 104 3.00 52 4.84 64 2.67 60 2.25 116 4.11 125 3.66 122
Ad-TV-NDC [36]104.4 8.75 127 15.3 98 12.3 137 10.5 133 24.2 132 12.3 133 8.96 127 28.2 124 11.5 119 5.31 126 22.8 117 5.55 127 4.03 68 5.79 99 2.86 57 2.80 73 10.0 60 2.87 78 3.04 54 4.52 38 2.66 59 4.62 137 5.79 137 30.9 145
AugFNG_ROB [144]105.3 7.79 122 28.5 132 5.02 110 9.56 130 18.5 119 11.3 132 8.78 126 26.2 121 17.3 132 3.46 115 9.97 67 2.86 112 5.26 126 6.24 121 4.77 116 4.44 112 14.5 106 4.15 108 2.99 51 5.22 77 1.30 3 1.85 101 3.18 94 2.10 95
Modified CLG [34]109.1 6.79 116 24.7 124 6.63 120 7.09 122 17.4 115 9.40 126 10.1 128 29.2 126 16.6 130 4.48 123 27.5 122 3.86 118 4.80 114 6.31 123 4.48 112 2.65 69 17.6 120 2.69 75 2.92 49 4.94 68 2.07 25 3.19 132 5.17 133 5.78 128
Filter Flow [19]109.2 6.76 115 17.6 103 4.37 100 5.01 112 17.6 116 5.49 112 5.98 114 26.3 122 18.4 134 7.23 129 29.9 125 6.91 130 5.12 122 6.23 119 5.36 128 5.23 125 11.9 84 4.95 121 6.64 139 8.75 142 3.75 91 0.95 31 2.15 53 1.21 55
LFNet_ROB [150]109.5 8.41 126 29.9 134 5.80 115 5.03 113 13.7 107 5.06 110 7.93 123 22.4 118 5.30 106 3.31 113 14.9 95 2.56 108 5.26 126 6.39 127 4.98 124 4.56 117 17.8 121 4.51 115 3.73 85 5.99 99 2.59 54 1.82 100 3.26 98 2.08 94
ResPWCR_ROB [145]110.2 8.23 124 22.9 121 6.93 121 4.20 107 12.3 100 4.43 105 5.27 112 15.1 96 5.64 109 3.74 117 17.2 109 3.36 116 4.79 113 5.55 86 5.34 127 4.99 123 13.6 101 5.09 124 4.72 113 6.59 116 2.89 66 2.38 121 3.59 110 2.92 109
IAOF2 [51]110.6 5.05 106 13.6 84 4.64 106 4.90 111 14.5 110 5.78 114 3.68 103 18.6 108 5.15 105 12.3 138 34.1 130 13.8 138 4.65 108 6.21 118 3.78 93 4.47 113 13.5 98 3.70 102 5.73 129 7.13 128 3.98 100 1.96 105 3.53 108 2.35 100
FlowNet2 [122]112.0 8.99 129 25.8 125 7.01 122 9.84 131 19.0 121 10.7 130 7.98 124 20.1 112 13.5 126 4.47 122 9.41 64 4.21 121 5.17 124 6.08 114 4.92 121 4.10 103 11.2 76 4.43 112 5.98 134 7.71 137 2.90 68 1.78 97 2.92 87 1.89 87
TVL1_ROB [139]112.0 13.5 137 26.7 129 16.1 141 14.6 137 23.8 131 16.6 138 16.9 136 36.5 137 25.4 139 11.9 137 33.8 127 12.8 136 4.66 109 6.23 119 3.96 98 2.38 58 16.0 112 2.89 80 2.32 18 4.57 41 1.49 8 5.63 141 6.44 138 12.6 140
EPMNet [133]112.1 8.88 128 26.2 128 7.20 125 9.32 129 18.2 117 10.0 128 7.14 117 18.8 110 12.3 122 4.79 124 12.3 87 4.62 125 5.17 124 6.08 114 4.92 121 4.10 103 11.2 76 4.43 112 4.88 119 7.05 125 2.56 51 1.93 104 3.57 109 2.07 93
BlockOverlap [61]112.4 6.80 117 12.1 75 5.94 117 5.51 117 13.6 104 6.58 117 5.32 113 22.2 117 7.30 113 4.20 118 16.7 106 4.06 120 4.45 98 5.39 73 5.11 126 4.91 120 12.5 88 4.34 110 6.77 141 7.13 128 9.52 140 2.02 106 3.24 97 9.49 136
HBpMotionGpu [43]113.2 5.92 111 15.0 96 4.79 109 7.78 124 22.4 128 9.04 125 7.17 118 39.2 141 17.3 132 3.31 113 13.4 92 3.14 115 4.71 111 5.88 104 4.84 118 3.74 97 13.5 98 3.54 96 5.96 133 7.17 130 3.68 86 2.24 115 3.43 102 4.65 124
2D-CLG [1]114.1 9.69 132 37.7 140 7.18 124 11.1 134 21.9 126 13.9 136 19.0 141 34.8 132 28.7 140 13.0 139 46.7 136 12.8 136 4.97 119 5.79 99 5.47 131 4.08 101 21.2 131 4.32 109 2.29 14 4.00 6 1.64 11 4.47 136 5.51 136 6.55 130
SPSA-learn [13]114.1 6.87 118 21.3 118 7.92 128 6.02 119 21.1 124 6.96 119 7.55 121 27.5 123 12.7 124 4.44 120 29.2 123 4.59 124 4.80 114 5.92 107 4.93 123 4.94 121 17.3 118 5.02 123 3.37 69 5.01 70 2.29 37 4.14 135 4.97 132 6.49 129
GraphCuts [14]114.8 6.34 112 17.1 101 5.55 114 5.30 115 20.9 122 5.26 111 6.05 116 20.4 113 12.4 123 2.85 108 20.9 112 2.15 100 4.74 112 5.95 108 4.90 120 8.69 137 12.6 89 5.19 126 5.79 130 6.40 110 6.80 130 2.45 124 3.48 104 3.59 119
GroupFlow [9]115.0 9.15 131 25.8 125 10.5 134 11.6 136 30.0 139 12.3 133 10.2 129 35.4 134 11.9 120 3.50 116 15.8 102 3.39 117 5.48 132 6.56 129 4.42 110 9.25 138 24.8 134 10.8 140 2.35 20 4.58 42 1.67 12 2.93 128 5.22 134 4.99 125
Black & Anandan [4]116.0 7.19 120 18.9 109 8.40 129 5.96 118 22.6 129 6.69 118 8.73 125 28.7 125 12.1 121 4.46 121 29.4 124 4.52 123 4.91 118 6.59 130 4.09 101 4.18 108 19.4 127 4.44 114 4.69 110 6.36 108 2.01 22 3.13 131 4.46 126 5.06 126
IAOF [50]117.0 6.54 114 18.3 106 7.13 123 6.99 121 18.4 118 7.90 123 7.71 122 32.3 129 8.44 115 8.21 132 31.8 126 9.78 134 4.61 105 6.01 111 4.14 103 4.35 111 18.9 126 3.43 93 4.69 110 6.08 103 3.31 77 3.23 133 4.69 129 15.9 143
2bit-BM-tele [98]120.0 8.99 129 18.6 108 10.2 132 4.45 109 11.3 94 5.04 108 4.66 110 17.3 101 4.41 102 5.23 125 21.7 115 5.04 126 4.99 120 5.96 109 5.46 130 6.47 131 18.3 122 7.49 132 7.77 143 8.57 141 12.5 143 2.29 118 3.89 119 2.99 112
Nguyen [33]122.3 8.16 123 23.0 122 7.57 127 16.5 140 22.7 130 19.3 140 16.8 135 36.0 135 20.7 137 13.8 140 39.5 132 14.7 140 5.40 129 6.44 128 6.70 135 4.54 116 18.5 124 5.42 127 3.50 76 5.02 71 2.08 26 4.00 134 5.50 135 8.53 134
UnFlow [129]122.5 19.4 145 44.0 145 10.2 132 11.3 135 21.2 125 12.4 135 18.1 138 36.0 135 15.5 128 7.47 131 34.0 129 6.40 129 7.10 140 7.11 134 8.76 140 8.31 136 24.8 134 9.26 136 5.13 121 6.50 113 1.52 9 1.57 85 3.14 93 2.01 91
Heeger++ [104]125.0 18.3 144 32.1 137 10.5 134 9.98 132 34.3 145 7.90 123 16.0 133 32.7 130 11.0 118 9.23 133 47.6 137 7.86 132 5.95 135 6.74 131 5.71 132 23.4 145 49.5 146 24.5 145 3.61 81 6.53 114 2.58 52 1.78 97 3.68 114 2.96 110
SILK [79]125.6 10.7 133 31.4 136 13.1 139 8.77 127 26.6 134 9.80 127 13.6 131 34.9 133 16.7 131 6.53 128 45.4 134 6.08 128 6.11 136 7.36 138 6.71 136 6.96 134 29.4 140 7.39 131 2.97 50 4.77 55 3.96 99 5.00 138 6.99 139 10.7 138
Horn & Schunck [3]126.6 8.40 125 27.2 130 9.62 130 7.28 123 28.3 136 7.55 122 13.3 130 31.9 128 15.8 129 7.35 130 48.5 139 7.69 131 5.84 134 7.34 137 5.45 129 5.80 129 25.8 137 6.79 130 5.25 123 7.11 127 2.11 28 5.21 139 8.30 141 6.66 131
Periodicity [78]130.4 12.3 135 51.4 153 7.39 126 9.23 128 38.3 147 11.1 131 34.7 147 48.1 147 36.0 146 4.27 119 57.7 146 3.99 119 24.4 147 73.2 147 16.2 146 29.6 147 74.3 147 29.4 147 3.29 63 5.67 91 1.90 18 6.64 142 44.8 147 21.5 144
FFV1MT [106]130.9 17.0 143 35.1 139 10.1 131 8.30 125 33.0 143 7.40 121 17.2 137 40.0 143 15.3 127 9.57 134 55.8 145 8.75 133 7.99 145 8.46 145 10.1 143 21.8 144 36.1 143 23.3 144 4.43 106 7.02 123 4.09 106 1.78 97 3.68 114 2.96 110
TI-DOFE [24]131.4 16.4 141 34.0 138 21.2 145 21.5 143 31.9 140 25.3 144 24.4 145 41.0 145 33.0 145 22.8 143 46.3 135 25.2 144 6.25 137 7.67 140 6.82 137 6.37 130 25.6 136 7.87 134 3.94 91 5.78 96 1.78 16 8.49 145 9.86 144 12.5 139
H+S_ROB [138]131.5 14.9 139 45.1 146 10.5 134 15.0 139 29.6 138 16.1 137 24.2 144 40.2 144 29.9 144 33.2 146 52.4 143 34.9 145 7.74 144 8.16 142 11.7 145 13.6 143 42.2 145 17.3 143 2.88 44 5.51 83 2.54 50 8.22 144 8.90 142 8.03 133
SLK [47]132.1 12.3 135 43.0 144 16.5 142 19.8 142 32.9 142 22.4 142 21.4 142 38.4 139 29.3 143 41.6 147 51.6 141 44.5 147 6.87 139 7.63 139 8.94 141 8.09 135 31.2 142 9.56 137 3.34 67 5.41 81 2.60 55 8.13 143 9.23 143 13.9 142
Adaptive flow [45]138.2 16.4 141 28.7 133 16.7 143 17.2 141 25.5 133 19.6 141 18.8 140 37.6 138 36.5 147 11.2 136 43.6 133 11.9 135 7.11 141 7.79 141 7.88 138 10.1 141 24.1 133 10.1 139 16.1 146 14.2 146 22.1 146 2.92 127 4.89 131 5.22 127
HCIC-L [99]139.0 24.1 147 31.3 135 12.9 138 27.6 146 28.7 137 69.9 147 16.0 133 30.6 127 23.1 138 18.1 142 55.4 144 17.8 142 7.39 142 8.35 144 8.32 139 12.6 142 18.3 122 14.4 142 25.6 147 23.8 147 23.6 147 2.62 125 4.51 127 9.36 135
PGAM+LK [55]139.8 14.0 138 40.6 142 18.8 144 14.9 138 33.1 144 17.8 139 14.4 132 32.9 131 19.3 135 15.7 141 63.6 147 14.9 141 6.36 138 6.81 132 9.14 142 9.83 140 30.7 141 9.83 138 10.4 144 12.2 145 10.3 141 5.30 140 7.26 140 13.0 141
FOLKI [16]140.2 11.0 134 41.0 143 14.5 140 24.9 145 32.3 141 36.7 145 18.7 139 43.8 146 20.5 136 10.9 135 50.5 140 13.8 138 7.42 143 8.28 143 10.6 144 9.75 139 36.9 144 12.1 141 4.77 118 7.29 132 11.0 142 12.2 146 11.4 145 36.4 146
Pyramid LK [2]143.1 15.8 140 28.2 131 30.4 147 35.8 147 28.0 135 49.6 146 22.3 143 38.6 140 29.1 142 31.8 145 51.7 142 39.0 146 18.3 146 24.8 146 24.1 147 26.7 146 28.6 139 26.7 146 7.19 142 8.98 143 7.70 136 32.7 147 40.6 146 57.0 147
AdaConv-v1 [126]148.1 45.2 148 50.1 147 46.0 148 78.3 148 79.6 148 76.9 148 78.3 148 73.0 148 77.6 148 79.2 148 80.8 148 79.6 148 83.7 148 84.3 148 83.6 148 82.1 148 80.6 148 81.5 148 69.6 149 58.5 149 75.3 149 84.4 148 84.7 148 83.9 148
SepConv-v1 [127]148.1 45.2 148 50.1 147 46.0 148 78.3 148 79.6 148 76.9 148 78.3 148 73.0 148 77.6 148 79.2 148 80.8 148 79.6 148 83.7 148 84.3 148 83.6 148 82.1 148 80.6 148 81.5 148 69.6 149 58.5 149 75.3 149 84.4 148 84.7 148 83.9 148
SuperSlomo [132]148.1 45.2 148 50.1 147 46.0 148 78.3 148 79.6 148 76.9 148 78.3 148 73.0 148 77.6 148 79.2 148 80.8 148 79.6 148 83.7 148 84.3 148 83.6 148 82.1 148 80.6 148 81.5 148 69.6 149 58.5 149 75.3 149 84.4 148 84.7 148 83.9 148
FGIK [136]148.1 45.2 148 50.1 147 46.0 148 78.3 148 79.6 148 76.9 148 78.3 148 73.0 148 77.6 148 79.2 148 80.8 148 79.6 148 83.7 148 84.3 148 83.6 148 82.1 148 80.6 148 81.5 148 69.6 149 58.5 149 75.3 149 84.4 148 84.7 148 83.9 148
CtxSyn [137]148.1 45.2 148 50.1 147 46.0 148 78.3 148 79.6 148 76.9 148 78.3 148 73.0 148 77.6 148 79.2 148 80.8 148 79.6 148 83.7 148 84.3 148 83.6 148 82.1 148 80.6 148 81.5 148 69.6 149 58.5 149 75.3 149 84.4 148 84.7 148 83.9 148
CyclicGen [154]148.1 45.2 148 50.1 147 46.0 148 78.3 148 79.6 148 76.9 148 78.3 148 73.0 148 77.6 148 79.2 148 80.8 148 79.6 148 83.7 148 84.3 148 83.6 148 82.1 148 80.6 148 81.5 148 69.6 149 58.5 149 75.3 149 84.4 148 84.7 148 83.9 148
AVG_FLOW_ROB [142]153.2 85.5 154 80.5 154 99.9 154 99.9 154 99.9 154 99.9 154 96.1 154 99.9 154 95.9 154 89.9 154 87.4 154 95.1 154 99.9 154 99.9 154 99.9 154 95.8 154 81.5 154 91.9 154 39.9 148 40.1 148 34.2 148 99.9 154 99.9 154 99.9 154
Move the mouse over the numbers in the table to see the corresponding images. Click to compare with the ground truth.

References

Methodtime*framescolor Reference and notes
[1] 2D-CLG 844 2 gray The 2D-CLG method by Bruhn et al. as implemented by Stefan Roth. [A. Bruhn, J. Weickert, and C. Schnörr. Lucas/Kanade meets Horn/Schunck: combining local and global optic flow methods. IJCV 63(3), 2005.] Parameters were set to match the published performance on Yosemite sequence, which may not be optimal for other sequences.
[2] Pyramid LK 12 2 color A modification of Bouguet's pyramidal implementation of Lucas-Kanade.
[3] Horn & Schunck 49 2 gray A modern Matlab implementation of the Horn & Schunck method by Deqing Sun. Parameters set to optimize AAE on all training data.
[4] Black & Anandan 328 2 gray A modern Matlab implementation of the Black & Anandan method by Deqing Sun.
[5] Brox et al. 18 2 color T. Brox, A. Bruhn, N. Papenberg, and J. Weickert. High accuracy optical flow estimation based on a theory for warping. ECCV 2004. (Improved using separate robust functions as proposed in A. Bruhn and J. Weickert, Towards ultimate motion estimation, ICCV 2005; improved by training on the training set.)
[6] Fusion 2,666 2 color V. Lempitsky, S. Roth, and C. Rother. Discrete-continuous optimization for optical flow estimation. CVPR 2008.
[7] Dynamic MRF 366 2 gray B. Glocker, N. Paragios, N. Komodakis, G. Tziritas, and N. Navab. Optical flow estimation with uncertainties through dynamic MRFs. CVPR 2008. (Method improved since publication.)
[8] Second-order prior 14 2 gray W. Trobin, T. Pock, D. Cremers, and H. Bischof. An unbiased second-order prior for high-accuracy motion estimation. DAGM 2008. (Method improved since publication; for details see W. Trobin, Ph.D. thesis, 2009.)
[9] GroupFlow 600 2 gray X. Ren. Local Grouping for Optical Flow. CVPR 2008.
[10] SegOF 60 2 color L. Xu, J. Chen, and J. Jia. Segmentation based variational model for accurate optical flow estimation. ECCV 2008. Code available.
[11] Learning Flow 825 2 gray D. Sun, S. Roth, J.P. Lewis, and M. Black. Learning optical flow (SRF-LFC). ECCV 2008.
[12] CBF 69 2 color W. Trobin, T. Pock, D. Cremers, and H. Bischof. Continuous energy minimization via repeated binary fusion. ECCV 2008. (Method improved since publication; for details see W. Trobin, Ph.D. thesis, 2009.)
[13] SPSA-learn 200 2 color Y. Li and D. Huttenlocher. Learning for optical flow using stochastic optimization. ECCV 2008.
[14] GraphCuts 1,200 2 color T. Cooke. Two applications of graph-cuts to image processing. DICTA 2008.
[15] F-TV-L1 8 2 gray A. Wedel, T. Pock, J. Braun, U. Franke, and D. Cremers. Duality TV-L1 flow with fundamental matrix prior. IVCNZ 2008.
[16] FOLKI 1.4 2 gray G. Le Besnerais and F. Champagnat. Dense optical flow by iterative local window registration. ICIP 2005.
[17] TV-L1-improved 2.9 2 gray A. Wedel, T. Pock, C. Zach, H. Bischof, and D. Cremers. An improved algorithm for TV-L1 optical flow computation. Proceedings of the Dagstuhl Visual Motion Analysis Workshop 2008. Code at GPU4Vision.
[18] DPOF 287 2 color C. Lei and Y.-H. Yang. Optical flow estimation on coarse-to-fine region-trees using discrete optimization. ICCV 2009. (Method improved since publication.)
[19] Filter Flow 34,000 2 color S. Seitz and S. Baker. Filter flow. ICCV 2009.
[20] Adaptive 9.2 2 gray A. Wedel, D. Cremers, T. Pock, and H. Bischof. Structure- and motion-adaptive regularization for high accuracy optic flow. ICCV 2009.
[21] Complementary OF 44 2 color H. Zimmer, A. Bruhn, J. Weickert, L. Valgaerts, A. Salgado, B. Rosenhahn, and H.-P. Seidel. Complementary optic flow. EMMCVPR 2009.
[22] Aniso. Huber-L1 2 2 gray M. Werlberger, W. Trobin, T. Pock, A. Wedel, D. Cremers, and H. Bischof. Anisotropic Huber-L1 optical flow. BMVC 2009. Code at GPU4Vision.
[23] Rannacher 0.12 2 gray J. Rannacher. Realtime 3D motion estimation on graphics hardware. Bachelor thesis, Heidelberg University, 2009.
[24] TI-DOFE 260 2 gray C. Cassisa, S. Simoens, and V. Prinet. Two-frame optical flow formulation in an unwarped multiresolution scheme. CIARP 2009.
[25] NL-TV-NCC 20 2 color M. Werlberger, T. Pock, and H. Bischof. Motion estimation with non-local total variation regularization. CVPR 2010.
[26] MDP-Flow 188 2 color L. Xu, J. Jia, and Y. Matsushita. Motion detail preserving optical flow estimation. CVPR 2010.
[27] ACK-Prior 5872 2 color K. Lee, D. Kwon, I. Yun, and S. Lee. Optical flow estimation with adaptive convolution kernel prior on discrete framework. CVPR 2010.
[28] LDOF 122 2 color T. Brox and J. Malik. Large displacement optical flow: descriptor matching in variational motion estimation. PAMI 33(3):500-513, 2011.
[29] p-harmonic 565 2 gray J. Gai and R. Stevenson. Optical flow estimation with p-harmonic regularization. ICIP 2010.
[30] TriangleFlow 4200 2 gray B. Glocker, H. Heibel, N. Navab, P. Kohli, and C. Rother. TriangleFlow: Optical flow with triangulation-based higher-order likelihoods. ECCV 2010.
[31] Classic+NL 972 2 color D. Sun, S. Roth, and M. Black. Secrets of optical flow estimation and their principles. CVPR 2010. Matlab code.
[32] Classic++ 486 2 gray A modern implementation of the classical formulation descended from Horn & Schunck and Black & Anandan; see D. Sun, S. Roth, and M. Black, Secrets of optical flow estimation and their principles, CVPR 2010.
[33] Nguyen 33 2 gray D. Nguyen. Tuning optical flow estimation with image-driven functions. ICRA 2011.
[34] Modified CLG 133 2 gray R. Fezzani, F. Champagnat, and G. Le Besnerais. Combined local global method for optic flow computation. EUSIPCO 2010.
[35] ComplOF-FED-GPU 0.97 2 color P. Gwosdek, H. Zimmer, S. Grewenig, A. Bruhn, and J. Weickert. A highly efficient GPU implementation for variational optic flow based on the Euler-Lagrange framework. CVGPU Workshop 2010.
[36] Ad-TV-NDC 35 2 gray M. Nawaz. Motion estimation with adaptive regularization and neighborhood dependent constraint. DICTA 2010.
[37] Layers++ 18206 2 color D. Sun, E. Sudderth, and M. Black. Layered image motion with explicit occlusions, temporal consistency, and depth ordering. NIPS 2010.
[38] OFH 620 3 color H. Zimmer, A. Bruhn, J. Weickert. Optic flow in harmony. IJCV 93(3) 2011.
[39] LSM 1615 2 color K. Jia, X. Wang, and X. Tang. Optical flow estimation using learned sparse model. ICCV 2011.
[40] CostFilter 55 2 color C. Rhemann, A. Hosni, M. Bleyer, C. Rother, and M. Gelautz. Fast cost-volume filtering for visual correspondence and beyond. CVPR 2011.
[41] Bartels 0.15 2 gray C. Bartels and G. de Haan. Smoothness constraints in recursive search motion estimation for picture rate conversion. IEEE TCSVT 2010. Version improved since publication: mapped on GPU.
[42] Shiralkar 600 2 gray M. Shiralkar and R. Schalkoff. A self organization-based optical flow estimator with GPU implementation. MVA 23(6):1229-1242.
[43] HBpMotionGpu 1000 5 gray S. Grauer-Gray and C. Kambhamettu. Hierarchical belief propagation to reduce search space using CUDA for stereo and motion estimation. WACV 2009. (Method improved since publication.)
[44] StereoFlow 7200 2 color G. Rosman, S. Shem-Tov, D. Bitton, T. Nir, G. Adiv, R. Kimmel, A. Feuer, and A. Bruckstein. Over-parameterized optical flow using a stereoscopic constraint. SSVM 2011:761-772.
[45] Adaptive flow 121 2 gray T. Arici. Energy minimization based motion estimation using adaptive smoothness priors. Submitted to IEEE TIP 2011.
[46] TC-Flow 2500 5 color S. Volz, A. Bruhn, L. Valgaerts, and H. Zimmer. Modeling temporal coherence for optical flow. ICCV 2011.
[47] SLK 300 2 gray T. Corpetti and E. Mémin. Stochastic uncertainty models for the luminance consistency assumption. IEEE TIP 2011.
[48] CLG-TV 29 2 gray M. Drulea. Total variation regularization of local-global optical flow. ITSC 2011. Matlab code.
[49] SimpleFlow 1.7 2 color M. Tao, J. Bai, P. Kohli, S. Paris. SimpleFlow: a non-iterative, sublinear optical flow algorithm. EUROGRAPHICS 2012.
[50] IAOF 57 2 gray D. Nguyen. Improving motion estimation using image-driven functions and hybrid scheme. PSIVT 2011.
[51] IAOF2 56 2 gray D. Nguyen. Enhancing the sharpness of flow field using image-driven functions with occlusion-aware filter. Submitted to TIP 2011.
[52] LocallyOriented 9541 2 gray Y.Niu, A. Dick, and M. Brooks. Locally oriented optical flow computation. To appear in TIP 2012.
[53] IROF-TV 261 2 color H. Rashwan, D. Puig, and M. Garcia. On improving the robustness of differential optical flow. ICCV 2011 Artemis workshop.
[54] Sparse Occlusion 2312 2 color A. Ayvaci, M. Raptis, and S. Soatto. Sparse occlusion detection with optical flow. Submitted to IJCV 2011.
[55] PGAM+LK 0.37 2 gray A. Alba, E. Arce-Santana, and M. Rivera. Optical flow estimation with prior models obtained from phase correlation. ISVC 2010.
[56] Sparse-NonSparse 713 2 color L. Chen, J. Wang, and Y. Wu. Decomposing and regularizing sparse/non-sparse components for motion field estimation. Submitted to PAMI 2013.
[57] nLayers 36150 4 color D. Sun, E. Sudderth, and M. Black. Layered segmentation and optical flow estimation over time. CVPR 2012.
[58] IROF++ 187 2 color H. Rashwan, D. Puig, and M. Garcia. Variational optical flow estimation based on stick tensor voting. IEEE TIP 2013.
[59] COFM 600 3 color M. Mozerov. Constrained optical flow estimation as a matching problem. IEEE TIP 2013.
[60] Efficient-NL 400 2 color P. Krähenbühl and V. Koltun. Efficient nonlocal regularization for optical flow. ECCV 2012.
[61] BlockOverlap 2 2 gray M. Santoro, G. AlRegib, and Y. Altunbasak. Motion estimation using block overlap minimization. Submitted to MMSP 2012.
[62] Ramp 1200 2 color A. Singh and N. Ahuja. Exploiting ramp structures for improving optical flow estimation. ICPR 2012.
[63] Occlusion-TV-L1 538 3 gray C. Ballester, L. Garrido, V. Lazcano, and V. Caselles. A TV-L1 optical flow method with occlusion detection. DAGM-OAGM 2012.
[64] TV-L1-MCT 90 2 color M. Mohamed and B. Mertsching. TV-L1 optical flow estimation with image details recovering based on modified census transform. ISVC 2012.
[65] Local-TV-L1 500 2 gray L. Raket. Local smoothness for global optical flow. ICIP 2012.
[66] ALD-Flow 61 2 color M. Stoll, A. Bruhn, and S. Volz. Adaptive integration of feature matches into variational optic flow methods. ACCV 2012.
[67] SIOF 234 2 color L. Xu, Z. Dai, and J. Jia. Scale invariant optical flow. ECCV 2012.
[68] MDP-Flow2 342 2 color L. Xu, J. Jia, and Y. Matsushita. Motion detail preserving optical flow estimation. PAMI 34(9):1744-1757, 2012. Code available.
[69] TCOF 1421 all gray J. Sanchez, A. Salgado, and N. Monzon. Optical flow estimation with consistent spatio-temporal coherence models. VISAPP 2013.
[70] LME 476 2 color W. Li, D. Cosker, M. Brown, and R. Tang. Optical flow estimation using Laplacian mesh energy. CVPR 2013.
[71] NN-field 362 2 color L. Chen, H. Jin, Z. Lin, S. Cohen, and Y. Wu. Large displacement optical flow from nearest neighbor fields. CVPR 2013.
[72] FESL 3310 2 color W. Dong, G. Shi, X. Hu, and Y. Ma. Nonlocal sparse and low-rank regularization for optical flow estimation. Submitted to IEEE TIP 2013.
[73] PMF 35 2 color J. Lu, H. Yang, D. Min, and M. Do. PatchMatch filter: efficient edge-aware filtering meets randomized search for fast correspondence field estimation. CVPR 2013.
[74] FC-2Layers-FF 2662 4 color D. Sun, J. Wulff, E. Sudderth, H. Pfister, and M. Black. A fully-connected layered model of foreground and background flow. CVPR 2013.
[75] Correlation Flow 290 2 color M. Drulea and S. Nedevschi. Motion estimation using the correlation transform. TIP 2013. Matlab code.
[76] TC/T-Flow 341 5 color M. Stoll, S. Volz, and A. Bruhn. Joint trilateral filtering for multiframe optical flow. ICIP 2013.
[77] OFLAF 1530 2 color T. Kim, H. Lee, and K. Lee. Optical flow via locally adaptive fusion of complementary data costs. ICCV 2013.
[78] Periodicity 8000 4 color G. Khachaturov, S. Gonzalez-Brambila, and J. Gonzalez-Trejo. Periodicity-based computation of optical flow. Submitted to Computacion y Sistemas (CyS) 2013.
[79] SILK 572 2 gray P. Zille, C. Xu, T. Corpetti, L. Shao. Observation models based on scale interactions for optical flow estimation. Submitted to IEEE TIP.
[80] CRTflow 13 3 color O. Demetz, D. Hafner, and J. Weickert. The complete rank transform: a tool for accurate and morphologically invariant matching of structures. BMVC 2013.
[81] SuperFlow 178 2 color Anonymous. Superpixel based optical flow estimation. ICCV 2013 submission 507.
[82] Aniso-Texture 300 2 color Anonymous. Texture information-based optical flow estimation using an incremental multi-resolution approach. ITC-CSCC 2013 submission 267.
[83] Classic+CPF 640 2 gray Z. Tu, R. Veltkamp, and N. van der Aa. A combined post-filtering method to improve accuracy of variational optical flow estimation. Submitted to Pattern Recognition 2013.
[84] S2D-Matching 1200 2 color Anonymous. Locally affine sparse-to-dense matching for motion and occlusion estimation. ICCV 2013 submission 1479.
[85] AGIF+OF 438 2 gray Z. Tu, R. Poppe, and R. Veltkamp. Adaptive guided image filter to warped interpolation image for variational optical flow computation. Submitted to Signal Processing 2015.
[86] DeepFlow 13 2 color P. Weinzaepfel, J. Revaud, Z. Harchaoui, and C. Schmid. DeepFlow: large displacement optical flow with deep matching. ICCV 2013.
[87] NNF-Local 673 2 color Z. Chen, H. Jin, Z. Lin, S. Cohen, and Y. Wu. Large displacement optical flow with nearest neighbor field. Submitted to PAMI 2014.
[88] EPPM w/o HM 2.5 2 color L. Bao, Q. Yang, and H. Jin. Fast edge-preserving PatchMatch for large displacement optical flow. CVPR 2014.
[89] MLDP_OF 165 2 gray M. Mohamed, H. Rashwan, B. Mertsching, M. Garcia, and D. Puig. Illumination-robust optical flow approach using local directional pattern. IEEE TCSVT 24(9):1499-1508, 2014.
[90] RFlow 20 2 gray S. Ali, C. Daul, and W. Blondel. Robust and accurate optical flow estimation for weak texture and varying illumination condition: Application to cystoscopy. IPTA 2014.
[91] SRR-TVOF-NL 32 all color P. Pohl, M. Sirotenko, E. Tolstaya, and V. Bucha. Edge preserving motion estimation with occlusions correction for assisted 2D to 3D conversion. IS&T/SPIE Electronic Imaging 2014.
[92] 2DHMM-SAS 157 2 color M.-C. Shih, R. Shenoy, and K. Rose. A two-dimensional hidden Markov model with spatially-adaptive states with application of optical flow. ICIP 2014 submission.
[93] WLIF-Flow 700 2 color Z. Tu, R. Veltkamp, N. van der Aa, and C. Van Gemeren. Weighted local intensity fusion method for variational optical flow estimation. Submitted to TIP 2014.
[94] FMOF 215 2 color N. Jith, A. Ramakanth, and V. Babu. Optical flow estimation using approximate nearest neighbor field fusion. ICASSP 2014.
[95] TriFlow 150 2 color TriFlow. Optical flow with geometric occlusion estimation and fusion of multiple frames. ECCV 2014 submission 914.
[96] ComponentFusion 6.5 2 color Anonymous. Fast optical flow by component fusion. ECCV 2014 submission 941.
[97] AggregFlow 1642 2 color D. Fortun, P. Bouthemy, and C. Kervrann. Aggregation of local parametric candidates and exemplar-based occlusion handling for optical flow. Preprint arXiv:1407.5759.
[98] 2bit-BM-tele 124 2 gray R. Xu and D. Taubman. Robust dense block-based motion estimation using a two-bit transform on a Laplacian pyramid. ICIP 2013.
[99] HCIC-L 330 2 color Anonymous. Globally-optimal image correspondence using a hierarchical graphical model. NIPS 2014 submission 114.
[100] TF+OM 600 2 color R. Kennedy and C. Taylor. Optical flow with geometric occlusion estimation and fusion of multiple frames. EMMCVPR 2015.
[101] PH-Flow 800 2 color J. Yang and H. Li. Dense, accurate optical flow estimation with piecewise parametric model. CVPR 2015.
[102] EpicFlow 16 2 color J. Revaud, P. Weinzaepfel, Z. Harchaoui, and C. Schmid. EpicFlow: edge-preserving interpolation of correspondences for optical flow. CVPR 2015.
[103] NNF-EAC 380 2 color Anonymous. Variational method for joint optical flow estimation and edge-aware image restoration. CVPR 2015 submission 2336.
[104] Heeger++ 6600 5 gray Anonymous. A context aware biologically inspired algorithm for optical flow (updated results). CVPR 2015 submission 2238.
[105] HBM-GC 330 2 color A. Zheng and Y. Yuan. Motion estimation via hierarchical block matching and graph cut. Submitted to ICIP 2015.
[106] FFV1MT 358 5 gray F. Solari, M. Chessa, N. Medathati, and P. Kornprobst. What can we expect from a V1-MT feedforward architecture for optical flow estimation? Submitted to Signal Processing: Image Communication 2015.
[107] ROF-ND 4 2 color S. Ali, C. Daul, E. Galbrun, and W. Blondel. Illumination invariant large displacement optical flow using robust neighbourhood descriptors. Submitted to CVIU 2015.
[108] DeepFlow2 16 2 color J. Revaud, P. Weinzaepfel, Z. Harchaoui, and C. Schmid. Deep convolutional matching. Submitted to IJCV, 2015.
[109] HAST 2667 2 color Anonymous. Highly accurate optical flow estimation on superpixel tree. ICCV 2015 submission 2221.
[110] FlowFields 15 2 color C. Bailer, B. Taetz, and D. Stricker. Flow Fields: Dense unregularized correspondence fields for highly accurate large displacement optical flow estimation. ICCV 2015.
[111] SVFilterOh 1.56 2 color Anonymous. Fast estimation of large displacement optical flow using PatchMatch and dominant motion patterns. CVPR 2016 submission 1788.
[112] FlowNetS+ft+v 0.5 2 color Anonymous. Learning optical flow with convolutional neural networks. ICCV 2015 submission 235.
[113] CombBMOF 51 2 color M. Brüggemann, R. Kays, P. Springer, and O. Erdler. Combined block-matching and adaptive differential motion estimation in a hierarchical multi-scale framework. ICGIP 2014. (Method improved since publication.)
[114] PMMST 182 2 color F. Zhang, S. Xu, and X. Zhang. High accuracy correspondence field estimation via MST based patch matching. Submitted to TIP 2015.
[115] DF-Auto 70 2 color N. Monzon, A. Salgado, and J. Sanchez. Regularization strategies for discontinuity-preserving optical flow methods. Submitted to TIP 2015.
[116] CPM-Flow 3 2 color Anonymous. Efficient coarse-to-fine PatchMatch for large displacement optical flow. CVPR 2016 submission 241.
[117] CNN-flow-warp+ref 1.4 3 color D. Teney and M. Hebert. Learning to extract motion from videos in convolutional neural networks. ArXiv 1601.07532, 2016.
[118] Steered-L1 804 2 color Anonymous. Optical flow estimation via steered-L1 norm. Submitted to WSCG 2016.
[119] StereoOF-V1MT 343 2 gray Anonymous. Visual features for action-oriented tasks: a cortical-like model for disparity and optic flow computation. BMVC 2016 submission 132.
[120] PGM-C 5 2 color Y. Li. Pyramidal gradient matching for optical flow estimation. Submitted to PAMI 2016.
[121] RNLOD-Flow 1040 2 gray C. Zhang, Z. Chen, M. Wang, M. Li, and S. Jiang. Robust non-local TV-L1 optical flow estimation with occlusion detection. IEEE TIP 26(8):4055-4067, 2017.
[122] FlowNet2 0.091 2 color Anonymous. FlowNet 2.0: Evolution of optical flow estimation with deep networks. CVPR 2017 submission 900.
[123] S2F-IF 20 2 color Anonymous. S2F-IF: Slow-to-fast interpolator flow. CVPR 2017 submission 765.
[124] BriefMatch 0.068 2 gray G. Eilertsen, P.-E. Forssen, and J. Unger. Dense binary feature matching for real-time optical flow estimation. SCIA 2017 submission 62.
[125] OAR-Flow 60 2 color Anonymous. Order-adaptive regularisation for variational optical flow: global, local and in between. SSVM 2017 submission 20.
[126] AdaConv-v1 2.8 2 color S. Niklaus, L. Mai, and F. Liu. (Interpolation results only.) Video frame interpolation via adaptive convolution. CVPR 2017.
[127] SepConv-v1 0.2 2 color S. Niklaus, L. Mai, and F. Liu. (Interpolation results only.) Video frame interpolation via adaptive separable convolution. ICCV 2017.
[128] ProbFlowFields 37 2 color A. Wannenwetsch, M. Keuper, and S. Roth. ProbFlow: joint optical flow and uncertainty estimation. ICCV 2017.
[129] UnFlow 0.12 2 color Anonymous. UnFlow: Unsupervised learning of optical flow with a bidirectional census loss. Submitted to AAAI 2018.
[130] FlowFields+ 10.5 2 color C. Bailer, B. Taetz, and D. Stricker. Flow fields: Dense correspondence fields for highly accurate large displacement optical flow estimation. Submitted to PAMI 2017.
[131] IIOF-NLDP 150 2 color D.-H. Trinh, W. Blondel, and C. Daul. A general form of illumination-invariant descriptors in variational optical flow estimation. ICIP 2017.
[132] SuperSlomo 0.5 2 color Anonymous. (Interpolation results only.) Super SloMo: High quality estimation of multiple intermediate frames for video interpolation. CVPR 2018 submission 325.
[133] EPMNet 0.061 2 color Anonymous. EPM-convolution multilayer-network for optical flow estimation. ICME 2018 submission 1119.
[134] OFRF 90 2 color T. Mai, M. Gouiffes, and S. Bouchafa. Optical Flow refinement using iterative propagation under color, proximity and flow reliability constraints. Submitted to Signal, Image and Video Processing 2017.
[135] 3DFlow 328 2 color J. Chen, Z. Cai, J. Lai, and X. Xie. A filtering based framework for optical flow estimation. IEEE TCSVT 2018.
[136] FGIK 0.18 2 color Anonymous. (Interpolation results only.) Learning flow-guided interpolation kernels for video frame synthesis. ECCV 2018 submission 433.
[137] CtxSyn 0.07 2 color S. Niklaus and F. Liu. (Interpolation results only.) Context-aware synthesis for video frame interpolation. CVPR 2018.
[138] H+S_ROB 5 2 color ROB 2018 baseline submission, based on: E. Meinhardt-Llopis, J. Sanchez, and D. Kondermann. Horn-Schunck optical flow with a multi-scale strategy. Image Processing On Line 3:151–172, 2013.
[139] TVL1_ROB 1 2 color ROB 2018 baseline submission, based on: J. Sanchez, E. Meinhardt-Llopis, and G. Facciolo. TV-L1 optical flow estimation. Image Processing On Line 3:137-150, 2013.
[140] DMF_ROB 10 2 color ROB 2018 baseline submission, based on: P. Weinzaepfel, J. Revaud, Z. Harchaoui, and C. Schmid. DeepFlow: large displacement optical flow with deep matching. ICCV 2013.
[141] JOF 657 2 gray C. Zhang, L. Ge, Z. Chen, M. Li, W. Liu, and H. Chen. Refined TV-L1 optical flow estimation using joint filtering. Submitted to IEEE TMM, 2018.
[142] AVG_FLOW_ROB N/A 2 N/A Average flow field of ROB 2018 training set.
[143] LiteFlowNet 0.06 2 color T.-W. Hui, X. Tang, and C. C. Loy. LiteFlowNet: A lightweight convolutional neural network for optical flow estimation. CVPR 2018.
[144] AugFNG_ROB 0.10 all color Anonymous. FusionNet and AugmentedFlowNet: Selective proxy ground truth for training on unlabeled images. ECCV 2018 submission 2834.
[145] ResPWCR_ROB 0.2 2 color Anonymous. Learning optical flow with residual connections. ROB 2018 submission.
[146] FF++_ROB 17.43 2 color R. Schuster, C. Bailer, O. Wasenmueller, D. Stricker. FlowFields++: Accurate optical flow correspondences meet robust interpolation. ICIP 2018. Submitted to ROB 2018.
[147] ProFlow_ROB 76 3 color Anonymous. ProFlow: Learning to predict optical flow. BMVC 2018 submission 277.
[148] PWC-Net_ROB 0.069 2 color D. Sun, X. Yang, M.-Y. Liu, and J. Kautz. PWC-Net: CNNs for optical flow using pyramid, warping, and cost volume. CVPR 2018.
[149] WOLF_ROB 0.02 2 color Anonymous. Reversed deep neural network for optical flow. ROB 2018 submission.
[150] LFNet_ROB 0.068 2 color Anonymous. Learning a flow network. ROB 2018 submission.
[151] WRT 9 2 color L. Mei, J. Lai, X. Xie, J. Zhu, and J. Chen. Illumination-invariance optical flow estimation using weighted regularization transform. Submitted to IEEE TCSVT 2018.
[152] EAI-Flow 2.1 2 color Anonymous. Hierarchical coherency sensitive hashing and interpolation with RANSAC for large displacement optical flow. CVIU 2018 submission 17-678.
[153] ContinualFlow_ROB 0.5 all color Anonymous. Continual Flow. ACCV 2018 submission 273.
[154] CyclicGen 0.088 2 color Anonymous. (Interpolation results only.) Deep video frame interpolation using cyclic frame generation. AAAI 2019 submission 323.
* 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.