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        
A95
angle
error
avg. Army
(Hidden texture)
GT   im0   im1
Mequon
(Hidden texture)
GT   im0   im1
Schefflera
(Hidden texture)
GT   im0   im1
Wooden
(Hidden texture)
GT   im0   im1
Grove
(Synthetic)
GT   im0   im1
Urban
(Synthetic)
GT   im0   im1
Yosemite
(Synthetic)
GT   im0   im1
Teddy
(Stereo)
GT   im0   im1
rank all disc untext all disc untext all disc untext all disc untext all disc untext all disc untext all disc untext all disc untext
NNF-Local [87]8.9 7.47 3 40.1 11 3.98 3 6.49 13 30.3 3 5.60 16 5.82 4 26.1 3 4.68 15 3.86 10 53.5 5 2.99 18 9.77 1 12.4 1 5.36 3 8.67 5 31.8 3 7.03 3 5.10 18 10.1 48 3.70 7 2.31 8 5.34 10 1.21 2
NN-field [71]12.2 8.38 14 43.1 26 4.19 4 7.34 22 28.7 2 6.26 26 5.82 4 28.9 6 4.68 15 2.94 2 54.1 6 2.16 3 10.4 4 13.2 4 5.24 2 6.12 1 17.5 1 4.46 1 6.24 46 10.6 63 4.10 8 2.35 12 6.44 19 1.14 1
MDP-Flow2 [68]14.0 8.02 8 38.6 6 5.75 18 5.17 2 31.1 4 4.55 3 5.48 3 30.8 9 4.22 8 4.49 19 99.9 43 3.27 26 11.3 13 13.4 6 8.04 18 10.8 14 54.4 32 10.5 21 4.84 7 9.33 31 4.31 11 2.69 26 4.85 5 2.20 3
OFLAF [77]15.7 7.70 5 39.8 8 4.74 8 6.40 12 32.5 7 5.82 21 4.73 2 25.3 2 3.96 4 4.47 18 99.9 43 3.55 42 10.2 3 13.0 2 6.29 7 13.3 32 42.1 19 9.90 18 5.10 18 8.01 10 4.66 21 2.75 27 5.59 14 6.33 33
PMMST [114]16.9 8.63 17 31.3 1 6.03 24 8.51 36 26.8 1 8.18 60 7.50 11 28.0 5 6.07 34 4.26 16 34.8 3 3.29 27 10.9 8 13.2 4 6.26 6 10.4 12 29.9 2 9.42 13 5.00 14 10.1 48 4.37 12 3.25 36 4.40 4 3.36 11
nLayers [57]23.9 8.19 11 45.3 52 4.62 6 9.65 59 31.7 6 8.88 77 8.87 18 33.6 11 8.22 61 3.62 6 99.9 43 2.93 15 10.5 5 13.6 8 6.52 8 11.3 20 33.4 4 9.45 15 6.02 42 8.56 16 4.99 29 2.31 8 6.80 24 5.53 30
NNF-EAC [103]24.1 8.80 18 40.8 14 6.14 26 6.13 7 39.3 26 5.36 11 6.97 9 35.1 12 4.73 17 5.83 38 87.9 29 3.49 38 12.1 26 14.6 20 8.87 29 12.5 29 41.2 15 11.8 34 5.35 24 10.1 48 4.61 19 3.19 34 7.62 34 3.99 21
ComponentFusion [96]25.2 8.30 13 49.1 71 5.87 21 5.69 5 35.4 15 5.40 12 7.24 10 35.3 13 4.99 21 3.69 7 99.9 43 2.32 5 11.7 19 14.1 13 8.75 27 15.8 54 66.6 53 15.0 68 5.71 32 8.88 21 5.09 34 2.64 23 5.25 9 3.68 16
FC-2Layers-FF [74]27.2 8.26 12 41.4 18 6.27 27 8.85 40 37.9 22 7.81 44 6.02 7 31.8 10 6.25 37 3.91 11 88.8 32 2.86 13 11.0 11 13.7 10 7.20 10 16.5 62 40.4 11 16.3 82 7.14 65 10.7 67 6.61 54 2.15 3 3.87 1 2.77 5
FESL [72]27.7 7.69 4 40.2 12 4.90 9 11.0 77 48.5 50 9.07 79 10.5 27 42.1 25 6.42 39 3.60 5 99.9 43 2.55 8 10.9 8 13.6 8 8.86 28 11.1 18 36.4 6 10.6 22 6.73 59 10.2 52 5.95 44 2.51 20 5.37 12 3.35 10
LME [70]28.4 7.85 7 42.7 24 6.01 23 5.33 3 34.6 10 4.91 7 14.6 49 54.5 46 40.7 98 4.66 21 73.0 12 3.25 23 11.5 16 13.8 11 9.65 51 11.6 22 70.4 62 12.1 39 5.14 20 9.97 46 4.51 18 2.86 28 6.45 20 4.29 26
TC/T-Flow [76]28.5 9.01 21 38.1 4 3.81 1 6.64 15 55.1 68 4.62 4 8.13 14 46.4 37 4.20 7 5.32 31 99.9 43 2.88 14 11.5 16 14.1 13 7.28 12 8.85 6 38.5 9 9.44 14 5.85 38 10.7 67 10.0 94 3.61 44 10.0 49 8.53 63
HAST [109]28.8 6.42 1 43.9 35 3.97 2 7.16 18 33.1 8 5.92 23 3.76 1 23.5 1 2.83 1 3.36 3 99.9 43 2.08 2 10.0 2 13.0 2 4.83 1 16.7 67 59.3 39 19.4 96 11.3 105 12.9 92 17.6 115 2.67 25 4.13 2 2.93 7
WLIF-Flow [93]30.6 8.04 9 40.7 13 5.53 15 7.98 31 34.6 10 7.20 38 8.75 17 40.5 20 5.74 31 4.27 17 96.8 41 2.94 16 13.4 74 16.1 82 9.77 55 13.5 36 41.3 17 11.8 34 5.65 30 9.12 27 5.96 45 2.29 7 7.30 27 6.96 42
PMF [73]31.2 9.38 28 48.9 69 4.67 7 7.10 17 37.5 19 5.58 15 7.91 12 30.0 8 4.06 5 4.89 27 99.9 43 3.29 27 10.6 7 14.0 12 5.76 5 12.4 28 54.8 33 11.5 31 11.2 104 17.9 123 11.1 98 2.06 1 4.95 8 4.00 22
ALD-Flow [66]31.3 8.18 10 41.6 22 4.51 5 6.32 10 54.8 66 5.03 9 10.7 31 61.8 53 4.24 9 4.24 15 99.9 43 2.61 9 11.7 19 14.3 15 7.21 11 10.8 14 61.4 41 9.93 19 5.96 41 9.55 36 9.77 93 3.87 48 15.7 58 9.38 75
Layers++ [37]32.0 9.07 22 44.4 38 8.41 62 8.47 35 31.5 5 8.03 53 5.85 6 37.9 15 6.02 33 3.76 9 62.6 7 2.79 11 10.5 5 13.5 7 8.22 23 17.6 75 55.0 34 14.5 65 7.44 70 10.9 71 5.70 43 2.27 6 4.86 6 9.14 68
Efficient-NL [60]33.0 8.40 15 43.6 32 5.38 13 9.12 44 37.4 18 7.83 46 11.1 35 56.6 48 6.16 35 5.71 35 99.9 43 3.32 30 11.3 13 14.7 23 7.66 13 16.6 66 37.6 7 12.7 47 6.93 63 10.8 69 5.98 47 2.89 30 5.47 13 2.90 6
RNLOD-Flow [121]33.0 7.26 2 38.7 7 5.25 12 7.65 27 43.0 33 6.20 25 12.7 42 75.2 59 4.59 13 3.52 4 99.9 43 2.46 6 11.4 15 14.9 31 7.74 15 16.2 56 40.2 10 16.2 79 8.20 78 12.2 84 7.63 71 2.49 19 5.98 17 7.10 45
SVFilterOh [111]33.9 8.96 20 55.2 97 5.63 16 7.20 19 37.8 21 6.39 29 6.18 8 43.0 28 5.02 22 3.71 8 99.9 43 2.49 7 11.1 12 14.3 15 5.54 4 13.5 36 50.6 31 14.1 60 10.1 96 16.8 120 12.4 107 2.20 4 4.91 7 2.50 4
TC-Flow [46]34.3 8.59 16 41.0 16 5.12 11 5.47 4 45.6 41 4.31 2 10.2 24 94.7 71 3.49 2 6.08 44 99.9 43 3.50 39 11.9 24 14.5 18 7.90 17 11.9 26 61.7 43 11.5 31 5.72 33 9.85 42 11.5 102 4.02 51 15.0 56 9.14 68
AGIF+OF [85]34.4 8.89 19 44.1 36 6.77 31 10.2 65 44.4 37 8.51 73 10.2 24 43.5 30 6.63 43 4.80 24 99.9 43 3.25 23 11.7 19 14.7 23 9.52 47 13.7 39 40.9 12 12.6 44 5.73 35 8.99 22 5.96 45 2.36 14 7.50 30 7.55 48
IROF++ [58]36.6 9.29 26 45.0 50 6.30 28 9.55 56 43.0 33 8.20 61 10.8 32 43.1 29 7.57 53 6.28 48 99.9 43 3.90 49 12.2 29 14.8 27 9.41 45 15.2 52 44.1 26 14.5 65 5.27 21 9.48 34 3.69 6 2.62 22 6.48 21 4.14 23
Correlation Flow [75]36.8 9.27 25 38.3 5 5.40 14 6.33 11 36.7 16 4.85 6 18.4 53 99.9 80 3.58 3 4.87 26 35.8 4 3.47 35 12.9 55 15.8 67 9.17 39 16.0 55 68.6 56 16.5 87 6.59 55 9.85 42 7.93 78 2.88 29 7.57 33 3.06 8
PH-Flow [101]37.0 10.2 37 44.6 42 7.86 44 9.41 51 42.0 32 8.11 56 8.41 16 38.9 17 7.40 52 6.39 50 99.9 43 3.93 51 11.8 22 14.5 18 7.72 14 13.8 40 42.9 23 13.1 50 7.21 66 10.3 55 7.61 70 2.34 11 4.20 3 4.20 24
Classic+CPF [83]37.2 9.64 31 43.4 28 7.93 47 9.43 52 46.1 44 7.82 45 10.6 30 51.0 41 6.68 44 5.09 29 99.9 43 3.22 22 12.0 25 15.2 42 9.15 37 14.7 47 34.4 5 13.4 53 6.42 50 10.1 48 6.89 59 2.26 5 6.77 23 7.08 44
CostFilter [40]40.0 10.5 43 46.8 61 6.98 35 7.51 24 38.1 23 6.31 28 9.22 19 29.7 7 4.74 18 5.86 40 99.9 43 3.97 53 10.9 8 14.3 15 6.77 9 13.4 33 56.1 35 12.3 41 11.6 108 20.5 127 14.2 109 2.12 2 8.52 40 6.71 39
IIOF-NLDP [131]41.4 12.0 56 43.5 30 5.86 20 9.35 47 35.0 12 6.81 32 11.2 36 88.4 67 4.51 11 6.09 45 28.8 2 4.09 57 14.6 99 18.4 112 9.71 52 14.6 46 66.2 52 14.7 67 4.90 12 9.54 35 4.82 27 3.09 32 7.73 36 3.18 9
OAR-Flow [125]41.5 11.1 49 48.8 68 5.85 19 9.88 61 82.9 102 6.47 30 27.3 70 99.9 80 8.06 57 6.75 56 99.9 43 2.80 12 12.4 35 15.1 40 8.20 22 10.3 10 58.1 37 8.37 9 4.07 2 8.06 12 5.45 40 4.81 59 9.74 48 6.36 34
ProbFlowFields [128]42.3 16.3 83 53.8 89 10.9 95 7.72 28 40.6 29 7.12 34 14.1 47 45.2 36 10.5 66 6.67 55 62.6 7 4.39 64 12.9 55 15.4 52 9.64 50 10.1 9 63.2 48 10.8 23 4.86 9 8.16 14 4.69 23 3.36 38 9.23 45 3.72 17
Sparse-NonSparse [56]42.4 9.96 32 44.2 37 8.85 70 9.39 50 50.6 55 8.08 55 10.1 23 43.7 32 7.21 48 6.10 46 88.2 30 3.41 33 12.5 40 15.5 60 8.96 31 16.3 58 41.9 18 16.2 79 6.48 52 9.05 24 6.27 50 2.33 10 7.33 28 7.76 56
MLDP_OF [89]42.7 11.8 55 41.7 23 8.40 61 6.97 16 35.2 14 5.88 22 11.3 39 65.3 55 5.23 25 4.76 23 99.9 43 3.09 19 12.2 29 14.8 27 8.87 29 13.4 33 48.1 28 17.2 92 10.0 95 10.9 71 18.1 116 3.58 43 8.07 39 4.53 28
LSM [39]43.0 10.0 35 42.9 25 8.48 63 9.36 49 49.6 53 7.99 50 10.5 27 43.6 31 6.80 45 5.80 36 88.6 31 3.38 32 12.5 40 15.4 52 9.03 33 16.5 62 42.3 21 16.3 82 6.94 64 9.84 39 6.71 55 2.42 17 7.96 38 7.72 53
Ramp [62]43.2 10.2 37 44.4 38 8.09 56 9.47 53 46.1 44 8.17 59 9.51 21 42.4 27 6.88 47 5.40 32 99.9 43 3.53 40 12.5 40 15.2 42 9.71 52 16.7 67 42.1 19 16.5 87 6.76 60 10.0 47 7.07 62 2.46 18 5.84 16 5.24 29
MDP-Flow [26]45.6 11.2 50 43.1 26 9.86 87 8.14 34 35.1 13 8.21 62 11.2 36 42.1 25 9.44 64 6.41 51 99.9 43 4.20 61 12.2 29 14.6 20 10.0 66 11.7 24 63.6 49 9.60 16 5.56 28 10.9 71 4.39 13 5.78 67 99.9 94 8.99 65
FMOF [94]46.1 9.17 23 43.6 32 8.04 53 10.0 63 48.1 47 8.48 71 8.35 15 38.3 16 6.49 41 5.08 28 99.9 43 3.45 34 12.6 46 15.4 52 9.19 40 18.1 81 41.2 15 15.5 73 6.67 58 10.6 63 7.47 68 3.00 31 16.1 59 7.74 54
Classic+NL [31]46.3 10.1 36 44.9 48 8.90 71 9.49 54 51.6 58 7.87 47 9.93 22 43.9 34 7.31 51 6.07 43 99.9 43 3.78 47 12.5 40 15.3 48 9.06 34 17.1 71 41.0 13 15.8 76 7.32 69 10.8 69 6.80 58 2.35 12 5.62 15 7.69 52
OFH [38]46.6 12.6 62 43.4 28 9.45 79 7.30 21 64.4 78 5.27 10 27.6 71 99.9 80 4.87 20 6.60 54 99.9 43 3.74 45 12.4 35 14.7 23 9.62 49 15.5 53 74.1 67 15.6 75 4.60 6 9.39 32 4.64 20 5.39 63 26.0 66 6.68 38
IROF-TV [53]47.0 10.4 41 44.5 41 8.16 57 9.69 60 51.1 57 8.44 66 12.6 41 46.8 38 7.27 50 6.80 57 87.5 28 3.93 51 13.0 61 15.7 64 10.4 71 18.3 83 86.9 88 13.7 57 4.44 5 7.40 5 3.05 4 2.60 21 7.55 32 7.56 49
TV-L1-MCT [64]47.0 9.57 29 44.7 44 8.66 66 10.9 76 48.1 47 9.11 80 11.8 40 58.1 50 6.61 42 4.74 22 99.9 43 3.34 31 12.9 55 15.2 42 9.89 60 17.8 79 47.8 27 16.0 78 5.28 22 8.09 13 7.71 72 3.33 37 7.26 26 7.53 47
NL-TV-NCC [25]47.1 10.7 45 40.8 14 6.45 29 8.52 37 41.1 30 6.30 27 11.2 36 93.6 70 4.18 6 5.99 42 75.9 15 4.02 54 13.2 67 16.2 88 10.1 68 16.7 67 70.9 63 16.3 82 6.56 54 9.91 45 7.05 61 4.76 57 16.9 60 3.56 14
CombBMOF [113]47.2 12.5 61 41.3 17 6.50 30 8.58 38 36.9 17 6.88 33 10.9 33 35.9 14 5.21 24 10.4 76 85.0 26 5.90 85 11.6 18 15.2 42 8.15 20 27.3 102 60.9 40 35.6 116 9.20 87 14.7 108 6.75 57 3.17 33 7.53 31 4.20 24
COFM [59]47.2 9.37 27 55.5 98 6.86 33 7.28 20 44.2 36 6.17 24 14.3 48 47.6 40 8.22 61 4.15 14 99.9 43 2.23 4 13.2 67 16.2 88 12.2 91 17.6 75 75.4 69 15.5 73 6.20 45 8.77 20 7.35 66 3.62 45 5.35 11 6.43 36
Adaptive [20]47.9 10.2 37 46.1 55 4.95 10 9.63 57 55.4 69 7.80 43 36.7 84 99.9 80 7.64 54 6.15 47 78.7 17 2.96 17 12.1 26 14.8 27 9.09 35 12.3 27 85.8 86 6.06 2 8.72 82 12.5 88 4.97 28 3.55 42 34.8 70 9.13 67
AggregFlow [97]49.5 13.2 63 62.1 116 6.79 32 14.9 87 73.1 87 10.6 87 26.8 69 55.1 47 20.5 89 5.48 33 99.9 43 3.67 43 12.5 40 15.0 34 7.76 16 8.55 4 38.2 8 8.90 11 5.78 36 10.6 63 4.74 25 5.43 64 8.53 41 7.58 50
RFlow [90]50.6 11.5 53 45.3 52 8.80 69 6.22 9 49.2 52 5.41 13 26.2 67 99.9 80 5.04 23 4.14 13 99.9 43 3.11 20 12.6 46 15.0 34 9.87 58 16.5 62 83.2 81 13.8 59 6.62 56 8.58 17 6.16 48 6.33 72 99.9 94 12.0 94
S2F-IF [123]50.6 20.0 95 51.4 78 9.91 88 9.64 58 48.3 49 7.93 48 19.7 55 41.7 23 13.6 78 9.98 73 84.3 23 5.40 78 12.8 53 15.3 48 9.92 63 10.9 17 62.4 46 10.9 25 5.00 14 10.4 58 5.30 38 3.71 46 8.55 42 3.87 19
Sparse Occlusion [54]51.0 9.98 34 41.5 21 7.82 43 9.00 43 40.5 28 8.28 64 13.5 45 85.5 64 5.96 32 5.82 37 99.9 43 3.90 49 13.0 61 15.9 68 9.77 55 13.8 40 49.9 30 12.3 41 13.6 119 15.7 116 7.81 75 3.51 39 9.05 43 6.42 35
Complementary OF [21]51.2 13.6 67 46.2 56 9.35 76 6.20 8 50.4 54 4.92 8 12.8 43 58.8 52 5.45 27 7.89 64 99.9 43 5.59 80 12.3 33 14.6 20 9.99 65 18.9 85 69.9 60 14.3 63 5.44 25 7.80 7 7.78 74 6.13 70 26.9 67 9.66 82
S2D-Matching [84]51.3 9.96 32 53.0 86 8.51 64 9.53 55 53.0 60 7.94 49 20.5 57 99.9 80 6.80 45 5.30 30 83.0 19 3.53 40 12.4 35 15.2 42 9.16 38 17.3 72 41.1 14 16.8 90 7.75 74 10.5 61 7.90 77 2.36 14 6.34 18 9.58 80
FlowFields+ [130]51.5 20.3 96 52.0 80 10.3 91 10.3 67 44.8 39 8.44 66 19.5 54 40.2 19 14.1 79 10.2 75 66.6 10 6.28 90 12.8 53 15.4 52 9.97 64 10.3 10 61.8 44 10.2 20 4.85 8 10.9 71 4.79 26 3.97 50 12.8 53 3.85 18
Occlusion-TV-L1 [63]52.3 10.4 41 44.9 48 6.90 34 8.77 39 53.3 62 7.54 40 33.8 80 99.9 80 7.96 55 5.88 41 99.9 43 3.48 37 13.6 82 16.3 90 10.6 74 9.50 7 80.1 75 8.60 10 6.12 44 8.69 19 4.39 13 6.52 75 99.9 94 9.37 72
2DHMM-SAS [92]52.6 10.3 40 44.8 46 8.03 52 10.5 72 52.4 59 8.21 62 21.6 58 97.4 75 8.20 60 6.88 58 99.9 43 3.86 48 12.4 35 15.0 34 9.87 58 17.7 77 43.3 25 15.9 77 6.81 61 10.2 52 7.15 65 2.65 24 7.68 35 7.29 46
FlowFields [110]52.6 20.3 96 52.3 82 10.2 89 10.2 65 49.0 51 8.46 69 20.3 56 40.5 20 14.7 80 10.8 80 76.6 16 6.16 88 12.9 55 15.4 52 10.0 66 11.1 18 69.9 60 11.0 27 4.97 13 8.63 18 5.12 35 4.04 53 14.0 54 3.98 20
HBM-GC [105]52.8 10.9 47 57.5 106 7.03 37 9.35 47 40.2 27 8.80 75 8.09 13 52.3 43 6.42 39 6.91 59 84.3 23 6.17 89 11.8 22 14.7 23 8.40 24 14.7 47 43.2 24 12.6 44 9.81 94 17.8 122 8.50 84 3.54 41 10.1 50 10.1 86
Aniso-Texture [82]53.2 7.75 6 40.0 10 5.87 21 7.36 23 41.5 31 7.19 37 41.3 92 99.9 80 6.22 36 2.73 1 65.4 9 1.92 1 12.9 55 15.3 48 10.1 68 29.2 104 99.9 101 16.5 87 11.8 111 14.9 112 8.22 81 3.80 47 12.6 52 8.58 64
SimpleFlow [49]53.6 11.3 52 46.6 58 9.79 86 10.7 74 45.0 40 9.15 81 23.1 61 99.9 80 8.38 63 8.00 66 99.9 43 3.72 44 12.7 50 15.5 60 9.36 44 16.3 58 42.6 22 15.3 72 5.91 40 9.61 37 5.39 39 2.39 16 7.08 25 9.41 76
ACK-Prior [27]53.8 10.7 45 37.9 3 7.90 46 6.01 6 38.5 24 4.80 5 10.2 24 41.5 22 4.35 10 4.56 20 99.9 43 3.75 46 13.2 67 15.9 68 11.3 83 27.3 102 82.2 78 23.1 103 11.6 108 14.9 112 16.2 113 6.43 73 15.5 57 6.11 32
EPPM w/o HM [88]54.8 15.3 79 41.4 18 8.08 55 7.60 25 33.9 9 5.66 18 13.0 44 47.0 39 5.57 28 8.73 70 99.9 43 4.81 73 12.6 46 15.7 64 10.8 77 18.6 84 62.9 47 16.4 85 11.9 114 12.5 88 17.2 114 3.20 35 7.49 29 6.00 31
PGM-C [120]55.6 20.6 99 54.2 92 9.59 83 10.1 64 60.8 73 8.47 70 22.3 60 44.3 35 14.8 82 10.7 79 99.9 43 4.15 59 13.1 65 15.4 52 9.90 61 11.8 25 64.6 51 12.0 38 4.86 9 7.96 8 5.01 31 4.78 58 14.4 55 7.01 43
ROF-ND [107]56.6 12.0 56 39.9 9 8.22 60 6.49 13 45.6 41 5.49 14 13.5 45 92.7 68 4.83 19 8.05 67 23.7 1 5.54 79 14.2 94 17.5 101 11.2 82 20.1 91 72.0 65 15.0 68 13.0 117 13.3 98 10.5 97 3.52 40 6.67 22 3.36 11
ComplOF-FED-GPU [35]57.0 13.2 63 44.7 44 7.95 48 9.18 45 82.6 101 5.63 17 15.3 50 58.5 51 5.67 29 7.59 62 99.9 43 4.68 69 12.3 33 14.8 27 9.20 41 18.2 82 83.8 82 16.4 85 7.54 72 9.84 39 11.1 98 5.44 65 31.6 69 7.74 54
Steered-L1 [118]58.8 9.19 24 36.3 2 6.06 25 4.59 1 39.2 25 4.30 1 9.30 20 52.3 43 4.57 12 4.86 25 99.9 43 3.30 29 13.6 82 15.9 68 12.3 92 24.7 101 77.0 71 20.2 97 15.1 122 13.7 103 40.0 123 14.7 104 91.5 92 20.9 106
CPM-Flow [116]59.6 20.6 99 54.3 93 9.58 81 10.3 67 62.5 76 8.50 72 21.9 59 43.8 33 14.7 80 10.6 78 99.9 43 4.06 55 13.2 67 15.4 52 9.85 57 12.6 30 68.9 57 13.3 51 5.02 16 9.16 29 5.04 33 5.27 61 19.2 61 9.63 81
EpicFlow [102]60.1 20.6 99 54.1 91 9.59 83 10.3 67 62.9 77 8.55 74 26.3 68 99.4 79 15.1 84 10.4 76 99.9 43 4.07 56 13.1 65 15.4 52 9.90 61 11.6 22 67.3 54 11.9 37 4.86 9 7.97 9 4.99 29 5.34 62 19.2 61 9.74 84
DeepFlow2 [108]60.2 14.3 72 47.4 64 7.03 37 10.4 70 77.1 92 8.01 51 23.3 62 99.9 80 11.8 72 16.1 88 99.9 43 4.41 65 12.4 35 15.0 34 8.16 21 13.4 33 68.4 55 14.2 61 5.65 30 9.07 26 8.50 84 8.52 88 92.9 93 10.6 89
SRR-TVOF-NL [91]60.3 14.4 75 46.7 59 8.18 59 13.1 82 74.0 89 8.44 66 24.1 64 63.2 54 11.9 73 6.51 53 85.1 27 3.25 23 12.1 26 15.0 34 10.3 70 17.5 73 61.6 42 13.4 53 10.4 100 12.3 86 8.92 87 5.52 66 7.83 37 7.58 50
TCOF [69]61.5 13.6 67 44.8 46 8.02 51 9.90 62 54.6 65 8.02 52 31.3 76 99.9 80 15.4 86 6.49 52 82.4 18 4.76 71 14.9 103 18.0 109 9.50 46 9.71 8 48.3 29 12.6 44 10.1 96 12.7 90 8.96 88 4.29 54 9.21 44 6.80 40
F-TV-L1 [15]62.8 15.6 80 47.4 64 13.4 100 18.8 96 99.1 110 11.6 88 43.1 96 99.9 80 11.3 71 14.7 86 99.9 43 7.03 93 12.2 29 14.9 31 9.00 32 13.5 36 99.9 101 7.56 6 6.41 49 10.5 61 4.23 9 3.91 49 80.3 84 3.38 13
DPOF [18]62.8 17.4 91 49.1 71 7.77 42 12.3 78 45.6 41 8.91 78 10.9 33 26.6 4 8.09 58 7.81 63 99.3 42 5.31 75 13.5 78 16.0 78 11.0 80 17.5 73 61.8 44 12.2 40 13.1 118 10.9 71 18.1 116 5.04 60 9.50 47 4.49 27
TF+OM [100]63.0 12.1 58 51.3 77 7.13 39 8.92 42 44.4 37 8.13 58 33.8 80 54.4 45 45.8 101 6.28 48 90.1 34 4.63 68 12.7 50 15.2 42 10.6 74 18.9 85 99.9 101 11.3 29 7.81 75 14.2 105 6.33 51 7.09 82 43.5 72 8.22 59
Aniso. Huber-L1 [22]63.5 11.7 54 43.8 34 8.16 57 13.6 84 66.3 80 12.0 90 35.9 82 99.9 80 10.5 66 10.0 74 72.9 11 5.00 74 13.4 74 16.3 90 9.61 48 15.1 51 63.7 50 7.96 7 8.96 85 11.6 79 7.95 79 4.02 51 26.9 67 7.97 58
TV-L1-improved [17]65.2 10.9 47 45.2 51 7.42 40 8.12 32 54.0 63 6.79 31 36.5 83 99.9 80 7.26 49 5.84 39 99.9 43 3.15 21 13.2 67 15.9 68 9.11 36 22.1 95 99.9 101 20.8 98 9.59 92 13.3 98 9.04 89 6.19 71 88.8 88 9.71 83
SIOF [67]66.5 10.6 44 49.7 75 7.01 36 14.8 86 85.9 104 8.40 65 49.7 104 98.3 76 49.2 104 12.0 83 99.9 43 5.88 83 13.5 78 15.9 68 10.8 77 16.3 58 74.2 68 13.6 56 5.51 27 9.02 23 4.42 16 6.52 75 19.5 63 9.85 85
CRTflow [80]67.9 15.0 77 46.3 57 7.89 45 8.87 41 54.9 67 7.15 35 30.1 73 99.9 80 8.03 56 9.30 72 99.9 43 4.50 67 13.0 61 15.7 64 8.05 19 32.5 108 99.9 101 34.3 115 6.62 56 9.72 38 7.52 69 9.30 92 99.9 94 14.7 99
Brox et al. [5]68.6 16.0 81 49.2 73 12.0 97 12.3 78 80.4 97 10.3 85 23.7 63 73.1 58 13.2 75 24.2 96 99.9 43 4.23 62 14.7 100 16.8 95 15.4 113 10.7 13 96.7 94 9.71 17 5.88 39 9.05 24 3.01 3 8.78 89 67.7 80 9.37 72
DeepFlow [86]68.6 14.7 76 49.0 70 9.78 85 12.9 80 79.3 96 9.80 83 30.1 73 96.1 74 24.4 91 21.4 95 99.9 43 5.36 77 12.5 40 15.1 40 8.59 25 14.0 43 71.9 64 15.1 71 5.46 26 8.01 10 8.73 86 14.2 102 99.9 94 15.7 103
LocallyOriented [52]69.6 17.0 90 55.7 99 8.00 50 17.0 93 82.3 100 12.1 91 42.4 95 99.9 80 14.8 82 9.13 71 89.2 33 4.79 72 13.4 74 16.1 82 9.20 41 10.8 14 58.1 37 11.8 34 6.89 62 10.6 63 7.14 64 7.78 86 74.9 81 9.42 77
BriefMatch [124]69.6 9.63 30 44.4 38 5.74 17 7.64 26 51.0 56 5.70 19 10.5 27 39.2 18 4.63 14 3.95 12 99.9 43 2.72 10 16.2 114 17.6 105 33.0 125 41.4 120 99.4 100 43.4 122 12.7 116 13.2 96 67.6 127 79.5 120 99.9 94 99.9 122
Dynamic MRF [7]71.0 14.0 71 50.7 76 9.58 81 7.75 29 85.7 103 5.76 20 31.5 77 99.9 80 5.23 25 7.97 65 99.9 43 4.10 58 13.0 61 15.6 63 10.7 76 30.4 107 99.9 101 29.5 112 5.64 29 7.52 6 9.61 92 67.3 118 99.9 94 66.7 118
Classic++ [32]71.5 11.2 50 49.4 74 9.13 72 9.34 46 68.4 82 8.11 56 30.7 75 95.1 73 10.2 65 5.59 34 99.9 43 3.47 35 13.5 78 16.1 82 10.4 71 19.7 89 99.9 101 17.6 93 8.38 79 11.5 78 8.30 83 7.20 84 99.9 94 9.54 79
SuperFlow [81]73.7 14.3 72 47.0 63 9.26 74 19.5 98 58.7 72 17.9 98 45.9 99 99.9 80 56.1 107 19.0 93 99.9 43 5.88 83 13.3 73 16.1 82 12.6 96 11.5 21 74.0 66 8.27 8 9.24 88 12.9 92 4.70 24 8.27 87 89.2 90 8.28 60
Rannacher [23]73.9 13.8 70 47.5 66 10.8 94 10.5 72 62.0 75 8.84 76 41.1 91 99.9 80 11.0 69 8.49 68 99.9 43 4.28 63 13.5 78 16.1 82 9.72 54 22.5 97 99.9 101 17.0 91 7.66 73 9.88 44 7.82 76 4.72 56 75.1 82 9.37 72
CBF [12]74.7 12.2 60 41.4 18 8.65 65 16.5 90 47.1 46 16.6 97 24.7 65 88.1 66 12.9 74 11.0 82 99.9 43 4.18 60 14.9 103 17.5 101 14.0 108 15.0 50 79.8 74 8.97 12 14.9 121 15.1 114 15.9 112 5.78 67 63.0 79 10.1 86
p-harmonic [29]75.1 15.1 78 48.5 67 14.1 102 10.4 70 53.1 61 9.31 82 41.9 93 99.9 80 15.1 84 19.4 94 99.9 43 10.6 101 12.7 50 14.9 31 11.8 86 18.0 80 85.6 85 18.4 95 7.85 76 10.4 58 5.46 41 6.98 81 99.9 94 9.28 71
Local-TV-L1 [65]75.5 16.5 85 52.3 82 11.7 96 27.7 104 96.9 107 22.5 103 68.8 111 99.9 80 47.5 102 34.6 105 99.9 43 7.04 94 12.6 46 15.0 34 9.25 43 17.7 77 84.7 83 13.7 57 5.09 17 7.36 4 5.03 32 20.7 107 88.8 88 29.4 111
CLG-TV [48]76.4 12.1 58 43.5 30 9.42 78 14.1 85 60.9 74 13.2 93 33.2 79 99.9 80 11.2 70 10.8 80 84.7 25 5.82 82 14.8 102 17.9 107 12.1 89 13.8 40 99.9 101 11.3 29 10.9 102 14.2 105 9.22 90 6.69 78 99.9 94 8.52 62
FlowNet2 [122]76.8 32.0 115 68.0 124 14.9 104 35.2 109 77.9 95 29.8 112 32.9 78 51.2 42 35.4 96 16.3 89 99.9 43 10.5 100 13.2 67 15.9 68 12.0 88 14.1 44 99.9 101 10.8 23 8.84 84 20.1 124 4.66 21 4.49 55 9.46 46 3.65 15
TriFlow [95]77.4 16.1 82 57.1 104 7.97 49 13.3 83 65.2 79 12.3 92 48.8 102 99.9 80 61.8 111 7.03 60 91.4 36 5.33 76 13.4 74 15.3 48 11.4 84 14.3 45 76.1 70 13.3 51 22.7 127 14.3 107 26.7 120 5.93 69 11.8 51 7.86 57
SegOF [10]77.6 22.8 108 54.8 94 15.4 106 27.9 105 56.0 70 27.4 108 39.3 86 87.9 65 33.2 95 37.5 106 75.4 14 22.3 106 14.4 97 16.3 90 14.7 111 21.7 93 99.9 101 24.5 105 4.09 3 7.28 2 2.18 2 6.79 79 48.3 75 6.93 41
DF-Auto [115]77.7 20.5 98 55.9 101 9.21 73 22.7 100 74.8 90 19.3 101 44.7 98 93.5 69 57.8 109 27.1 99 99.9 43 5.70 81 15.0 106 18.9 119 11.8 86 7.13 2 57.9 36 7.09 4 10.2 99 13.9 104 4.23 9 8.94 90 54.0 77 9.21 70
Bartels [41]79.0 13.3 66 55.0 95 10.2 89 8.13 33 43.2 35 7.67 41 18.1 52 69.0 57 6.30 38 8.49 68 99.9 43 6.05 87 13.9 88 16.1 82 13.9 106 21.8 94 99.9 101 21.5 101 10.6 101 13.5 100 20.3 118 12.3 97 99.9 94 26.9 109
TriangleFlow [30]79.2 13.2 63 46.8 61 9.41 77 10.8 75 73.2 88 7.30 39 26.1 66 99.9 80 5.70 30 7.23 61 99.9 43 4.46 66 17.0 119 21.3 124 15.2 112 23.0 98 69.8 59 22.9 102 9.71 93 16.1 117 9.40 91 6.89 80 23.8 64 11.5 92
Fusion [6]80.7 16.3 83 53.8 89 12.5 98 7.93 30 37.7 20 7.75 42 15.6 51 41.8 24 13.2 75 13.5 84 83.1 20 7.77 97 15.4 108 18.5 113 14.2 110 33.1 109 89.0 89 24.8 107 11.8 111 14.7 108 8.27 82 11.4 96 99.9 94 13.3 96
CNN-flow-warp+ref [117]81.8 21.3 104 57.3 105 15.1 105 16.8 92 54.4 64 15.9 96 41.0 90 99.9 80 28.8 92 28.6 101 99.9 43 7.45 96 14.0 91 15.9 68 14.0 108 16.5 62 84.8 84 10.9 25 5.30 23 8.23 15 8.13 80 99.9 123 99.9 94 99.9 122
StereoFlow [44]82.5 48.0 126 74.6 130 41.1 126 61.0 122 99.9 111 51.6 121 71.4 112 99.9 80 63.9 113 65.6 120 99.9 43 61.2 120 16.2 114 15.9 68 22.6 120 7.22 3 77.8 73 7.39 5 3.38 1 7.35 3 1.99 1 7.18 83 99.9 94 11.4 91
LDOF [28]84.0 17.9 93 53.5 88 8.72 67 18.7 95 92.6 106 11.8 89 29.5 72 67.1 56 20.9 90 29.0 102 99.9 43 8.92 99 14.2 94 16.4 93 13.7 103 18.9 85 97.5 98 15.0 68 6.26 48 10.3 55 10.1 95 10.4 94 99.9 94 10.3 88
StereoOF-V1MT [119]84.3 16.5 85 46.0 54 9.27 75 18.9 97 99.9 111 7.16 36 40.5 88 99.9 80 8.18 59 14.7 86 96.4 40 6.30 91 14.1 93 16.8 95 12.9 98 29.9 105 91.4 91 27.0 110 5.78 36 10.4 58 10.4 96 99.9 123 99.9 94 99.9 122
FlowNetS+ft+v [112]84.9 16.5 85 52.1 81 8.06 54 17.4 94 76.9 91 13.4 94 46.3 100 99.9 80 30.4 94 29.8 103 99.9 43 13.8 103 15.5 109 18.5 113 13.8 105 12.7 31 89.5 90 11.7 33 9.24 88 13.5 100 12.1 105 7.39 85 57.9 78 9.52 78
Shiralkar [42]85.3 16.8 88 44.6 42 9.46 80 16.5 90 98.8 109 8.05 54 42.0 94 99.9 80 10.8 68 18.4 92 99.9 43 8.02 98 12.9 55 15.5 60 10.4 71 30.2 106 99.9 101 25.1 108 11.4 107 11.8 82 15.8 111 22.2 109 99.9 94 17.5 105
Learning Flow [11]85.4 13.7 69 52.8 84 7.67 41 12.9 80 87.1 105 10.0 84 40.5 88 95.0 72 13.4 77 38.1 107 99.9 43 4.74 70 17.1 121 21.7 125 12.5 95 24.2 100 99.9 101 13.5 55 7.95 77 12.7 90 6.98 60 23.9 111 99.9 94 14.9 100
Ad-TV-NDC [36]85.8 31.0 114 53.3 87 33.1 123 70.2 123 99.9 111 49.0 120 93.2 123 99.9 80 54.0 106 38.9 108 95.0 38 29.4 109 13.8 87 17.3 100 8.71 26 14.7 47 77.1 72 13.0 49 6.24 46 9.84 39 5.19 37 46.4 116 76.4 83 54.0 116
Second-order prior [8]87.7 14.3 72 46.7 59 8.79 68 15.2 88 72.5 86 10.5 86 39.2 85 99.9 80 16.6 88 17.5 90 99.9 43 6.01 86 14.4 97 17.5 101 10.8 77 38.6 118 99.9 101 24.7 106 11.3 105 12.2 84 11.2 100 9.13 91 89.5 91 15.6 102
Filter Flow [19]87.7 21.6 106 57.7 107 14.4 103 24.6 102 77.5 94 18.1 99 54.3 106 80.8 60 66.3 117 52.8 113 91.0 35 46.5 114 13.6 82 16.0 78 12.3 92 17.0 70 69.6 58 14.2 61 12.0 115 16.1 117 7.39 67 6.58 77 37.5 71 8.36 61
2D-CLG [1]89.6 46.1 124 67.5 123 28.2 119 39.5 113 77.3 93 38.9 115 93.9 125 99.9 80 74.9 122 53.6 114 99.9 43 51.0 116 13.9 88 15.9 68 13.5 102 24.0 99 99.9 101 21.2 99 4.28 4 7.24 1 4.50 17 12.7 98 99.9 94 11.6 93
GraphCuts [14]89.9 21.7 107 52.8 84 10.4 92 39.2 112 99.9 111 23.1 104 39.7 87 58.0 49 49.6 105 25.6 98 74.6 13 7.31 95 13.6 82 15.9 68 13.2 101 37.8 115 97.6 99 16.2 79 9.36 90 11.4 76 11.7 103 10.3 93 99.9 94 15.0 101
HBpMotionGpu [43]90.6 19.5 94 63.6 117 13.5 101 31.0 106 99.9 111 27.4 108 99.9 128 99.9 80 59.3 110 18.2 91 99.9 43 6.69 92 13.6 82 16.0 78 12.4 94 16.2 56 91.5 92 11.1 28 11.1 103 13.0 95 6.73 56 21.3 108 99.9 94 21.6 107
SPSA-learn [13]91.1 23.6 109 55.7 99 20.1 111 32.9 108 99.9 111 25.2 105 91.2 122 99.9 80 64.5 115 49.6 111 99.9 43 31.2 110 14.0 91 16.0 78 13.0 100 19.9 90 99.9 101 23.3 104 6.53 53 9.13 28 4.40 15 15.8 105 99.9 94 16.5 104
IAOF2 [51]91.4 16.8 88 59.5 112 10.5 93 20.1 99 69.0 84 18.1 99 53.3 105 99.9 80 56.8 108 55.3 117 95.0 38 54.7 118 14.2 94 17.2 99 11.0 80 19.2 88 81.1 76 14.3 63 11.8 111 13.2 96 13.0 108 13.5 99 45.0 73 9.00 66
Modified CLG [34]92.8 27.9 112 58.6 110 23.4 114 26.5 103 71.1 85 26.2 107 93.4 124 99.9 80 73.1 120 49.2 110 99.9 43 22.6 107 15.2 107 18.1 110 13.7 103 16.3 58 99.9 101 12.8 48 6.43 51 10.2 52 11.7 103 11.3 95 99.9 94 11.3 90
UnFlow [129]93.9 49.8 127 69.9 127 23.3 113 32.8 107 57.8 71 33.0 113 49.4 103 98.6 77 39.8 97 53.6 114 99.9 43 52.1 117 16.2 114 18.5 113 20.6 117 35.3 112 99.9 101 38.4 118 8.81 83 11.4 76 3.08 5 6.46 74 99.9 94 6.47 37
IAOF [50]94.2 20.6 99 55.0 95 17.4 108 36.6 110 99.9 111 27.6 110 99.9 128 99.9 80 75.5 123 32.7 104 93.3 37 25.5 108 13.9 88 16.6 94 12.1 89 36.5 113 92.3 93 12.3 41 9.40 91 11.7 80 7.13 63 26.2 112 47.4 74 28.8 110
GroupFlow [9]96.4 27.3 111 66.8 121 21.6 112 41.1 114 99.9 111 35.1 114 71.4 112 99.9 80 61.8 111 25.1 97 99.9 43 14.4 104 14.7 100 17.9 107 11.7 85 40.6 119 97.2 97 40.6 120 5.72 33 11.7 80 6.48 52 16.4 106 53.8 76 23.0 108
Black & Anandan [4]96.7 21.5 105 51.7 79 19.8 110 38.6 111 99.9 111 25.8 106 81.3 115 99.9 80 65.5 116 50.4 112 99.9 43 31.6 111 15.5 109 19.1 120 12.8 97 22.4 96 96.8 96 18.2 94 10.1 96 12.4 87 5.16 36 13.9 101 99.9 94 12.9 95
Nguyen [33]97.1 27.2 110 56.2 102 18.7 109 46.7 117 99.9 111 44.1 119 97.7 127 99.9 80 74.1 121 45.7 109 99.9 43 38.0 112 16.4 117 17.7 106 21.8 119 20.8 92 99.9 101 21.2 99 7.21 66 9.42 33 5.61 42 14.6 103 99.9 94 14.2 98
2bit-BM-tele [98]101.1 20.8 103 59.0 111 16.3 107 16.4 89 68.5 83 15.3 95 43.9 97 99.9 80 15.4 86 14.6 85 99.9 43 11.2 102 15.6 111 17.5 101 16.8 115 34.9 111 99.9 101 31.1 113 18.8 125 20.4 126 30.8 121 23.7 110 99.9 94 61.9 117
BlockOverlap [61]102.5 17.6 92 56.5 103 13.1 99 24.0 101 66.9 81 21.5 102 67.6 110 99.9 80 49.0 103 28.2 100 99.9 43 15.2 105 17.6 123 18.3 111 32.7 124 38.1 116 81.6 77 26.6 109 14.5 120 15.5 115 67.6 127 39.8 114 82.5 85 84.7 119
Heeger++ [104]102.9 33.0 116 58.3 109 23.6 115 60.7 120 99.9 111 39.4 116 59.1 107 99.9 80 30.1 93 86.5 129 99.9 43 70.5 122 14.9 103 17.1 97 12.9 98 76.5 127 99.9 101 79.8 127 7.49 71 12.9 92 6.59 53 99.9 123 99.9 94 99.9 122
SILK [79]105.0 33.3 117 64.6 118 29.2 121 46.3 116 99.9 111 39.4 116 95.0 126 99.9 80 66.5 118 56.7 118 99.9 43 50.2 115 16.1 113 18.7 117 17.2 116 48.7 122 99.9 101 37.3 117 7.26 68 9.22 30 14.5 110 70.9 119 99.9 94 51.7 115
Horn & Schunck [3]105.4 28.8 113 57.7 107 25.3 116 41.1 114 99.9 111 28.2 111 80.0 114 99.9 80 75.9 124 75.5 121 99.9 43 66.3 121 15.7 112 18.5 113 13.9 106 41.9 121 99.9 101 41.4 121 11.6 108 13.6 102 6.16 48 45.4 115 99.9 94 39.5 113
TI-DOFE [24]106.7 40.2 120 61.2 115 38.6 125 60.2 119 99.9 111 53.7 122 90.4 121 99.9 80 78.2 126 83.9 127 99.9 43 82.8 128 16.6 118 19.4 122 16.4 114 38.1 116 99.9 101 38.7 119 8.51 81 10.3 55 7.75 73 56.3 117 99.9 94 49.7 114
HCIC-L [99]109.0 44.4 123 66.7 120 29.1 120 99.9 129 99.9 111 99.9 129 47.8 101 99.9 80 44.6 100 58.5 119 99.9 43 55.7 119 20.3 125 20.6 123 24.0 122 33.9 110 86.5 87 33.6 114 38.0 129 49.5 129 36.0 122 13.7 100 25.8 65 13.7 97
FFV1MT [106]110.8 36.4 119 71.0 128 25.8 117 50.3 118 99.9 111 39.7 118 65.6 109 99.1 78 41.6 99 99.9 130 99.9 43 97.4 131 20.2 124 19.3 121 33.1 126 62.5 124 99.9 101 71.1 126 9.04 86 14.8 110 11.2 100 99.9 123 99.9 94 99.9 122
SLK [47]111.0 53.1 128 66.3 119 59.5 131 60.9 121 98.1 108 58.6 123 89.7 119 99.9 80 67.7 119 99.9 130 99.9 43 95.1 130 17.0 119 18.8 118 23.5 121 56.6 123 99.9 101 51.2 123 8.43 80 11.9 83 12.2 106 99.9 123 99.9 94 99.9 122
PGAM+LK [55]112.6 43.6 121 67.1 122 43.7 127 73.8 124 99.9 111 77.5 124 61.9 108 82.9 63 63.9 113 76.6 122 99.9 43 72.5 123 17.1 121 17.1 97 31.6 123 66.2 125 99.9 101 64.8 125 18.9 126 20.3 125 22.1 119 99.9 123 99.9 94 99.9 122
Adaptive flow [45]112.8 34.4 118 59.7 113 29.7 122 84.5 127 99.9 111 77.7 125 87.6 118 99.9 80 92.8 130 54.9 116 99.9 43 39.3 113 20.6 126 23.8 127 20.9 118 37.5 114 96.7 94 29.2 111 35.2 128 30.1 128 58.6 126 38.1 113 99.9 94 33.0 112
AdaConv-v1 [126]114.3 66.7 130 69.5 125 54.6 128 80.0 125 82.0 98 79.5 126 81.7 116 82.0 61 80.9 127 82.3 124 83.1 20 82.6 126 85.8 128 85.9 128 85.2 128 88.2 129 83.0 79 82.9 129 75.7 130 67.1 130 77.8 130 85.1 121 86.4 86 84.8 120
SepConv-v1 [127]114.3 66.7 130 69.5 125 54.6 128 80.0 125 82.0 98 79.5 126 81.7 116 82.0 61 80.9 127 82.3 124 83.1 20 82.6 126 85.8 128 85.9 128 85.2 128 88.2 129 83.0 79 82.9 129 75.7 130 67.1 130 77.8 130 85.1 121 86.4 86 84.8 120
Periodicity [78]114.7 54.4 129 84.3 131 26.5 118 99.9 129 99.9 111 99.9 129 99.9 128 99.9 80 99.9 131 81.4 123 99.9 43 76.3 124 99.9 131 99.9 130 99.9 131 99.9 131 99.9 101 99.9 131 6.06 43 14.8 110 70.6 129 99.9 123 99.9 94 99.9 122
FOLKI [16]116.8 43.8 122 74.4 129 38.0 124 99.9 129 99.9 111 99.9 129 89.7 119 99.9 80 76.2 125 85.0 128 99.9 43 81.0 125 23.2 127 22.5 126 38.8 127 66.2 125 99.9 101 62.7 124 17.0 124 17.6 121 42.6 124 99.9 123 99.9 94 99.9 122
Pyramid LK [2]117.7 47.7 125 60.7 114 56.1 130 88.6 128 99.9 111 91.9 128 99.9 128 99.9 80 89.4 129 83.0 126 99.9 43 83.2 129 98.4 130 99.9 130 87.9 130 87.4 128 99.9 101 81.4 128 16.9 123 16.6 119 57.0 125 99.9 123 99.9 94 99.9 122
Move the mouse over the numbers in the table to see the corresponding images. Click to compare with the ground truth.

References

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