Optical flow evaluation results |
Statistics:
Average
SD
R0.5
R1.0
R2.0
A50
A75
A95
Error type: endpoint angle interpolation normalized interpolation |
A50 endpoint error |
avg. |
Army (Hidden texture) GT im0 im1 |
Mequon (Hidden texture) GT im0 im1 |
Schefflera (Hidden texture) GT im0 im1 |
Wooden (Hidden texture) GT im0 im1 |
Grove (Synthetic) GT im0 im1 |
Urban (Synthetic) GT im0 im1 |
Yosemite (Synthetic) GT im0 im1 |
Teddy (Stereo) GT im0 im1 | ||||||||||||||||
rank | all | disc | untext | all | disc | untext | all | disc | untext | all | disc | untext | all | disc | untext | all | disc | untext | all | disc | untext | all | disc | untext | |
ComplexFlow [81] | 8.8 | 0.04 4 | 0.06 2 | 0.03 1 | 0.07 3 | 0.22 10 | 0.07 3 | 0.10 1 | 0.11 6 | 0.11 1 | 0.05 15 | 0.13 9 | 0.04 5 | 0.14 8 | 0.23 6 | 0.11 10 | 0.08 26 | 0.15 9 | 0.09 39 | 0.07 14 | 0.08 2 | 0.14 23 | 0.14 1 | 0.27 10 | 0.12 2 |
NN-field [73] | 9.0 | 0.04 4 | 0.06 2 | 0.04 4 | 0.08 8 | 0.25 19 | 0.08 7 | 0.10 1 | 0.11 6 | 0.12 2 | 0.04 5 | 0.13 9 | 0.04 5 | 0.13 4 | 0.22 4 | 0.10 8 | 0.07 10 | 0.13 2 | 0.07 8 | 0.09 41 | 0.09 12 | 0.17 35 | 0.14 1 | 0.28 19 | 0.11 1 |
Epistemic [84] | 9.4 | 0.03 1 | 0.07 7 | 0.03 1 | 0.07 3 | 0.20 5 | 0.07 3 | 0.11 4 | 0.11 6 | 0.12 2 | 0.03 1 | 0.17 29 | 0.03 1 | 0.14 8 | 0.25 10 | 0.10 8 | 0.07 10 | 0.28 50 | 0.06 4 | 0.07 14 | 0.10 32 | 0.10 9 | 0.16 4 | 0.27 10 | 0.15 3 |
ALD-Flow [68] | 10.1 | 0.04 4 | 0.07 7 | 0.04 4 | 0.07 3 | 0.19 2 | 0.07 3 | 0.11 4 | 0.11 6 | 0.12 2 | 0.04 5 | 0.12 5 | 0.04 5 | 0.13 4 | 0.25 10 | 0.09 3 | 0.05 2 | 0.16 13 | 0.06 4 | 0.06 4 | 0.08 2 | 0.22 56 | 0.18 22 | 0.34 32 | 0.20 40 |
TC/T-Flow [80] | 10.1 | 0.03 1 | 0.07 7 | 0.03 1 | 0.06 1 | 0.25 19 | 0.06 1 | 0.11 4 | 0.12 21 | 0.12 2 | 0.03 1 | 0.12 5 | 0.03 1 | 0.13 4 | 0.25 10 | 0.08 1 | 0.05 2 | 0.17 16 | 0.05 2 | 0.05 2 | 0.07 1 | 0.22 56 | 0.18 22 | 0.32 28 | 0.19 34 |
ADF [67] | 10.4 | 0.04 4 | 0.09 29 | 0.04 4 | 0.08 8 | 0.22 10 | 0.08 7 | 0.11 4 | 0.11 6 | 0.13 13 | 0.03 1 | 0.16 22 | 0.03 1 | 0.15 12 | 0.27 19 | 0.09 3 | 0.06 4 | 0.16 13 | 0.06 4 | 0.08 27 | 0.09 12 | 0.16 29 | 0.17 8 | 0.26 4 | 0.16 6 |
TC-Flow [46] | 11.0 | 0.04 4 | 0.08 20 | 0.04 4 | 0.06 1 | 0.18 1 | 0.06 1 | 0.11 4 | 0.11 6 | 0.12 2 | 0.04 5 | 0.10 3 | 0.04 5 | 0.13 4 | 0.26 14 | 0.09 3 | 0.06 4 | 0.20 25 | 0.07 8 | 0.05 2 | 0.08 2 | 0.20 51 | 0.18 22 | 0.34 32 | 0.20 40 |
MDP-Flow2 [70] | 11.2 | 0.05 26 | 0.08 20 | 0.05 16 | 0.07 3 | 0.19 2 | 0.07 3 | 0.11 4 | 0.11 6 | 0.12 2 | 0.05 15 | 0.12 5 | 0.05 29 | 0.15 12 | 0.23 6 | 0.11 10 | 0.08 26 | 0.14 4 | 0.07 8 | 0.07 14 | 0.09 12 | 0.12 16 | 0.18 22 | 0.25 1 | 0.16 6 |
nLayers [57] | 11.2 | 0.04 4 | 0.05 1 | 0.04 4 | 0.11 48 | 0.24 13 | 0.12 53 | 0.11 4 | 0.10 2 | 0.13 13 | 0.04 5 | 0.08 1 | 0.04 5 | 0.12 3 | 0.18 2 | 0.09 3 | 0.07 10 | 0.12 1 | 0.06 4 | 0.08 27 | 0.09 12 | 0.15 25 | 0.17 8 | 0.27 10 | 0.17 11 |
OFLADF [82] | 12.7 | 0.05 26 | 0.07 7 | 0.05 16 | 0.08 8 | 0.20 5 | 0.08 7 | 0.10 1 | 0.09 1 | 0.12 2 | 0.05 15 | 0.11 4 | 0.04 5 | 0.11 1 | 0.17 1 | 0.09 3 | 0.09 39 | 0.15 9 | 0.08 25 | 0.07 14 | 0.08 2 | 0.18 41 | 0.19 34 | 0.26 4 | 0.19 34 |
Layers++ [37] | 13.3 | 0.04 4 | 0.07 7 | 0.05 16 | 0.10 36 | 0.24 13 | 0.11 46 | 0.11 4 | 0.10 2 | 0.13 13 | 0.04 5 | 0.08 1 | 0.04 5 | 0.11 1 | 0.18 2 | 0.08 1 | 0.07 10 | 0.14 4 | 0.07 8 | 0.09 41 | 0.10 32 | 0.19 49 | 0.17 8 | 0.25 1 | 0.17 11 |
LME [72] | 14.0 | 0.05 26 | 0.07 7 | 0.04 4 | 0.08 8 | 0.19 2 | 0.08 7 | 0.11 4 | 0.12 21 | 0.13 13 | 0.05 15 | 0.18 34 | 0.05 29 | 0.16 19 | 0.26 14 | 0.11 10 | 0.07 10 | 0.18 22 | 0.07 8 | 0.07 14 | 0.09 12 | 0.13 20 | 0.18 22 | 0.27 10 | 0.16 6 |
IROF++ [58] | 14.2 | 0.04 4 | 0.07 7 | 0.05 16 | 0.09 17 | 0.30 38 | 0.10 28 | 0.11 4 | 0.12 21 | 0.13 13 | 0.05 15 | 0.16 22 | 0.04 5 | 0.15 12 | 0.26 14 | 0.12 14 | 0.07 10 | 0.21 27 | 0.07 8 | 0.07 14 | 0.09 12 | 0.09 3 | 0.17 8 | 0.28 19 | 0.17 11 |
Efficient-NL [60] | 14.8 | 0.04 4 | 0.06 2 | 0.04 4 | 0.09 17 | 0.30 38 | 0.09 18 | 0.11 4 | 0.11 6 | 0.13 13 | 0.04 5 | 0.14 12 | 0.04 5 | 0.14 8 | 0.23 6 | 0.11 10 | 0.07 10 | 0.16 13 | 0.07 8 | 0.09 41 | 0.09 12 | 0.20 51 | 0.18 22 | 0.28 19 | 0.18 26 |
Sparse-NonSparse [56] | 16.1 | 0.04 4 | 0.08 20 | 0.05 16 | 0.09 17 | 0.28 28 | 0.10 28 | 0.11 4 | 0.12 21 | 0.13 13 | 0.04 5 | 0.17 29 | 0.04 5 | 0.16 19 | 0.28 21 | 0.12 14 | 0.07 10 | 0.17 16 | 0.07 8 | 0.08 27 | 0.08 2 | 0.22 56 | 0.17 8 | 0.26 4 | 0.17 11 |
COFM [59] | 16.5 | 0.03 1 | 0.07 7 | 0.04 4 | 0.07 3 | 0.21 7 | 0.08 7 | 0.11 4 | 0.10 2 | 0.13 13 | 0.03 1 | 0.13 9 | 0.03 1 | 0.17 23 | 0.28 21 | 0.19 51 | 0.07 10 | 0.14 4 | 0.07 8 | 0.06 4 | 0.09 12 | 0.22 56 | 0.23 52 | 0.36 36 | 0.29 59 |
Levin3 [90] | 16.9 | 0.04 4 | 0.07 7 | 0.04 4 | 0.09 17 | 0.30 38 | 0.10 28 | 0.11 4 | 0.11 6 | 0.13 13 | 0.05 15 | 0.15 16 | 0.04 5 | 0.15 12 | 0.24 9 | 0.12 14 | 0.08 26 | 0.17 16 | 0.08 25 | 0.09 41 | 0.09 12 | 0.23 65 | 0.17 8 | 0.27 10 | 0.17 11 |
LSM [39] | 17.7 | 0.04 4 | 0.07 7 | 0.05 16 | 0.09 17 | 0.28 28 | 0.10 28 | 0.11 4 | 0.11 6 | 0.13 13 | 0.05 15 | 0.16 22 | 0.04 5 | 0.15 12 | 0.27 19 | 0.13 21 | 0.08 26 | 0.17 16 | 0.08 25 | 0.09 41 | 0.09 12 | 0.23 65 | 0.17 8 | 0.26 4 | 0.17 11 |
SCR [74] | 17.7 | 0.04 4 | 0.06 2 | 0.05 16 | 0.09 17 | 0.31 42 | 0.10 28 | 0.11 4 | 0.11 6 | 0.14 34 | 0.05 15 | 0.18 34 | 0.04 5 | 0.15 12 | 0.25 10 | 0.13 21 | 0.07 10 | 0.15 9 | 0.07 8 | 0.09 41 | 0.10 32 | 0.22 56 | 0.16 4 | 0.26 4 | 0.17 11 |
FC-2Layers-FF [77] | 17.8 | 0.04 4 | 0.07 7 | 0.05 16 | 0.09 17 | 0.27 25 | 0.10 28 | 0.11 4 | 0.10 2 | 0.14 34 | 0.05 15 | 0.12 5 | 0.04 5 | 0.14 8 | 0.22 4 | 0.12 14 | 0.08 26 | 0.15 9 | 0.08 25 | 0.10 60 | 0.10 32 | 0.23 65 | 0.17 8 | 0.26 4 | 0.17 11 |
Ramp [62] | 18.4 | 0.04 4 | 0.08 20 | 0.05 16 | 0.09 17 | 0.28 28 | 0.10 28 | 0.11 4 | 0.11 6 | 0.14 34 | 0.05 15 | 0.16 22 | 0.04 5 | 0.16 19 | 0.28 21 | 0.13 21 | 0.08 26 | 0.17 16 | 0.08 25 | 0.08 27 | 0.08 2 | 0.22 56 | 0.17 8 | 0.27 10 | 0.17 11 |
Classic+NL [31] | 20.5 | 0.04 4 | 0.07 7 | 0.05 16 | 0.09 17 | 0.29 34 | 0.10 28 | 0.11 4 | 0.11 6 | 0.14 34 | 0.05 15 | 0.15 16 | 0.04 5 | 0.15 12 | 0.26 14 | 0.13 21 | 0.08 26 | 0.18 22 | 0.08 25 | 0.10 60 | 0.10 32 | 0.23 65 | 0.17 8 | 0.27 10 | 0.17 11 |
TV-L1-MCT [64] | 20.9 | 0.04 4 | 0.07 7 | 0.04 4 | 0.10 36 | 0.33 45 | 0.11 46 | 0.11 4 | 0.11 6 | 0.14 34 | 0.05 15 | 0.16 22 | 0.04 5 | 0.17 23 | 0.29 27 | 0.16 39 | 0.08 26 | 0.21 27 | 0.08 25 | 0.07 14 | 0.09 12 | 0.11 13 | 0.18 22 | 0.28 19 | 0.18 26 |
SimpleFlow [49] | 22.3 | 0.04 4 | 0.08 20 | 0.05 16 | 0.10 36 | 0.31 42 | 0.11 46 | 0.12 30 | 0.13 26 | 0.14 34 | 0.05 15 | 0.17 29 | 0.04 5 | 0.16 19 | 0.28 21 | 0.15 33 | 0.08 26 | 0.17 16 | 0.08 25 | 0.08 27 | 0.08 2 | 0.17 35 | 0.17 8 | 0.27 10 | 0.17 11 |
FESL [75] | 22.5 | 0.04 4 | 0.06 2 | 0.05 16 | 0.10 36 | 0.34 49 | 0.10 28 | 0.11 4 | 0.11 6 | 0.13 13 | 0.05 15 | 0.14 12 | 0.05 29 | 0.17 23 | 0.26 14 | 0.16 39 | 0.08 26 | 0.14 4 | 0.08 25 | 0.09 41 | 0.11 51 | 0.20 51 | 0.17 8 | 0.28 19 | 0.18 26 |
Classic++ [32] | 24.5 | 0.04 4 | 0.08 20 | 0.05 16 | 0.09 17 | 0.25 19 | 0.10 28 | 0.11 4 | 0.13 26 | 0.14 34 | 0.05 15 | 0.17 29 | 0.04 5 | 0.17 23 | 0.35 35 | 0.13 21 | 0.08 26 | 0.25 38 | 0.08 25 | 0.09 41 | 0.10 32 | 0.24 70 | 0.18 22 | 0.32 28 | 0.17 11 |
Occlusion-TV-L1 [63] | 25.0 | 0.05 26 | 0.09 29 | 0.05 16 | 0.09 17 | 0.27 25 | 0.09 18 | 0.12 30 | 0.15 34 | 0.13 13 | 0.05 15 | 0.18 34 | 0.05 29 | 0.21 39 | 0.37 42 | 0.19 51 | 0.06 4 | 0.20 25 | 0.07 8 | 0.07 14 | 0.09 12 | 0.09 3 | 0.19 34 | 0.43 48 | 0.19 34 |
Correlation Flow [79] | 26.2 | 0.05 26 | 0.09 29 | 0.06 37 | 0.08 8 | 0.24 13 | 0.08 7 | 0.11 4 | 0.13 26 | 0.12 2 | 0.05 15 | 0.17 29 | 0.05 29 | 0.17 23 | 0.28 21 | 0.12 14 | 0.11 52 | 0.25 38 | 0.11 50 | 0.08 27 | 0.09 12 | 0.25 73 | 0.19 34 | 0.29 26 | 0.19 34 |
Adaptive [20] | 26.2 | 0.04 4 | 0.08 20 | 0.04 4 | 0.09 17 | 0.29 34 | 0.09 18 | 0.12 30 | 0.15 34 | 0.13 13 | 0.05 15 | 0.19 38 | 0.04 5 | 0.32 75 | 0.47 63 | 0.28 68 | 0.06 4 | 0.19 24 | 0.05 2 | 0.09 41 | 0.11 51 | 0.18 41 | 0.16 4 | 0.28 19 | 0.16 6 |
MDP-Flow [26] | 26.8 | 0.05 26 | 0.11 40 | 0.06 37 | 0.09 17 | 0.24 13 | 0.10 28 | 0.12 30 | 0.14 30 | 0.13 13 | 0.05 15 | 0.21 46 | 0.05 29 | 0.19 32 | 0.32 30 | 0.16 39 | 0.07 10 | 0.25 38 | 0.07 8 | 0.07 14 | 0.10 32 | 0.11 13 | 0.19 34 | 0.40 42 | 0.18 26 |
Direct ZNCC [66] | 26.9 | 0.05 26 | 0.09 29 | 0.06 37 | 0.08 8 | 0.25 19 | 0.08 7 | 0.11 4 | 0.14 30 | 0.12 2 | 0.05 15 | 0.16 22 | 0.05 29 | 0.19 32 | 0.31 29 | 0.13 21 | 0.11 52 | 0.27 46 | 0.11 50 | 0.08 27 | 0.09 12 | 0.25 73 | 0.18 22 | 0.28 19 | 0.19 34 |
IROF-TV [53] | 27.6 | 0.05 26 | 0.08 20 | 0.05 16 | 0.10 36 | 0.32 44 | 0.10 28 | 0.11 4 | 0.12 21 | 0.14 34 | 0.06 40 | 0.21 46 | 0.05 29 | 0.24 55 | 0.35 35 | 0.22 58 | 0.10 45 | 0.28 50 | 0.09 39 | 0.06 4 | 0.08 2 | 0.08 2 | 0.17 8 | 0.27 10 | 0.17 11 |
OFH [38] | 28.5 | 0.06 41 | 0.10 34 | 0.07 50 | 0.08 8 | 0.21 7 | 0.08 7 | 0.11 4 | 0.15 34 | 0.12 2 | 0.04 5 | 0.14 12 | 0.04 5 | 0.19 32 | 0.37 42 | 0.16 39 | 0.09 39 | 0.32 55 | 0.10 46 | 0.06 4 | 0.11 51 | 0.16 29 | 0.20 42 | 0.44 51 | 0.22 44 |
PMF [76] | 34.0 | 0.05 26 | 0.08 20 | 0.05 16 | 0.10 36 | 0.25 19 | 0.10 28 | 0.13 44 | 0.14 30 | 0.15 51 | 0.06 40 | 0.14 12 | 0.06 44 | 0.17 23 | 0.29 27 | 0.13 21 | 0.09 39 | 0.30 54 | 0.09 39 | 0.14 79 | 0.15 81 | 0.30 79 | 0.16 4 | 0.25 1 | 0.15 3 |
TV-L1-improved [17] | 34.5 | 0.04 4 | 0.09 29 | 0.04 4 | 0.08 8 | 0.24 13 | 0.08 7 | 0.12 30 | 0.15 34 | 0.13 13 | 0.05 15 | 0.18 34 | 0.04 5 | 0.22 45 | 0.40 49 | 0.13 21 | 0.15 72 | 0.41 69 | 0.17 78 | 0.11 69 | 0.13 69 | 0.25 73 | 0.18 22 | 0.38 39 | 0.18 26 |
Sparse Occlusion [54] | 38.7 | 0.06 41 | 0.10 34 | 0.05 16 | 0.11 48 | 0.28 28 | 0.12 53 | 0.12 30 | 0.15 34 | 0.13 13 | 0.06 40 | 0.19 38 | 0.05 29 | 0.21 39 | 0.33 32 | 0.15 33 | 0.11 52 | 0.22 30 | 0.09 39 | 0.17 87 | 0.16 84 | 0.22 56 | 0.19 34 | 0.32 28 | 0.17 11 |
Second-order prior [8] | 40.6 | 0.05 26 | 0.13 49 | 0.06 37 | 0.09 17 | 0.34 49 | 0.08 7 | 0.13 44 | 0.27 66 | 0.14 34 | 0.04 5 | 0.15 16 | 0.04 5 | 0.25 58 | 0.49 68 | 0.13 21 | 0.10 45 | 0.58 79 | 0.08 25 | 0.12 75 | 0.12 62 | 0.25 73 | 0.19 34 | 0.47 54 | 0.18 26 |
TriangleFlow [30] | 41.5 | 0.05 26 | 0.10 34 | 0.06 37 | 0.10 36 | 0.33 45 | 0.09 18 | 0.13 44 | 0.19 47 | 0.13 13 | 0.04 5 | 0.15 16 | 0.04 5 | 0.31 72 | 0.58 75 | 0.22 58 | 0.14 68 | 0.37 62 | 0.15 71 | 0.08 27 | 0.13 69 | 0.16 29 | 0.21 46 | 0.42 45 | 0.24 48 |
LDOF [28] | 42.0 | 0.06 41 | 0.14 53 | 0.07 50 | 0.10 36 | 0.36 56 | 0.10 28 | 0.13 44 | 0.25 64 | 0.14 34 | 0.06 40 | 0.33 59 | 0.05 29 | 0.22 45 | 0.41 52 | 0.22 58 | 0.07 10 | 0.25 38 | 0.07 8 | 0.07 14 | 0.10 32 | 0.13 20 | 0.26 64 | 0.62 69 | 0.35 64 |
CostFilter [40] | 42.2 | 0.06 41 | 0.10 34 | 0.06 37 | 0.11 48 | 0.27 25 | 0.11 46 | 0.13 44 | 0.16 40 | 0.15 51 | 0.08 57 | 0.15 16 | 0.09 62 | 0.18 29 | 0.32 30 | 0.13 21 | 0.10 45 | 0.36 60 | 0.10 46 | 0.14 79 | 0.17 86 | 0.33 83 | 0.15 3 | 0.32 28 | 0.15 3 |
Rannacher [23] | 42.4 | 0.06 41 | 0.12 45 | 0.07 50 | 0.11 48 | 0.30 38 | 0.11 46 | 0.13 44 | 0.19 47 | 0.15 51 | 0.06 40 | 0.27 54 | 0.05 29 | 0.22 45 | 0.41 52 | 0.15 33 | 0.11 52 | 0.34 57 | 0.10 46 | 0.09 41 | 0.10 32 | 0.18 41 | 0.18 22 | 0.37 37 | 0.18 26 |
Complementary OF [21] | 42.4 | 0.07 57 | 0.15 56 | 0.08 59 | 0.09 17 | 0.21 7 | 0.09 18 | 0.12 30 | 0.16 40 | 0.14 34 | 0.08 57 | 0.16 22 | 0.08 57 | 0.20 36 | 0.38 44 | 0.17 47 | 0.11 52 | 0.33 56 | 0.11 50 | 0.07 14 | 0.10 32 | 0.18 41 | 0.26 64 | 0.56 64 | 0.35 64 |
ComplOF-FED-GPU [35] | 42.6 | 0.07 57 | 0.15 56 | 0.08 59 | 0.08 8 | 0.26 24 | 0.08 7 | 0.12 30 | 0.18 44 | 0.12 2 | 0.07 50 | 0.15 16 | 0.07 54 | 0.22 45 | 0.45 60 | 0.16 39 | 0.12 61 | 0.40 66 | 0.12 58 | 0.09 41 | 0.10 32 | 0.23 65 | 0.21 46 | 0.47 54 | 0.24 48 |
ACK-Prior [27] | 43.2 | 0.07 57 | 0.11 40 | 0.07 50 | 0.09 17 | 0.22 10 | 0.09 18 | 0.12 30 | 0.13 26 | 0.13 13 | 0.07 50 | 0.19 38 | 0.06 44 | 0.20 36 | 0.34 34 | 0.14 32 | 0.13 66 | 0.29 53 | 0.12 58 | 0.11 69 | 0.11 51 | 0.42 87 | 0.25 58 | 0.40 42 | 0.28 57 |
Deep-Matching [85] | 43.5 | 0.07 57 | 0.15 56 | 0.11 69 | 0.12 56 | 0.33 45 | 0.13 58 | 0.13 44 | 0.23 60 | 0.17 59 | 0.08 57 | 0.28 55 | 0.08 57 | 0.18 29 | 0.35 35 | 0.12 14 | 0.06 4 | 0.22 30 | 0.07 8 | 0.06 4 | 0.08 2 | 0.18 41 | 0.29 70 | 0.60 66 | 0.38 67 |
F-TV-L1 [15] | 43.6 | 0.09 68 | 0.17 67 | 0.12 74 | 0.12 56 | 0.34 49 | 0.13 58 | 0.12 30 | 0.18 44 | 0.14 34 | 0.08 57 | 0.24 51 | 0.08 57 | 0.26 60 | 0.42 56 | 0.21 55 | 0.08 26 | 0.25 38 | 0.08 25 | 0.08 27 | 0.10 32 | 0.18 41 | 0.17 8 | 0.31 27 | 0.16 6 |
TCOF [71] | 43.7 | 0.06 41 | 0.12 45 | 0.07 50 | 0.12 56 | 0.34 49 | 0.12 53 | 0.14 60 | 0.21 55 | 0.16 55 | 0.09 64 | 0.19 38 | 0.10 65 | 0.24 55 | 0.45 60 | 0.12 14 | 0.07 10 | 0.13 2 | 0.08 25 | 0.11 69 | 0.12 62 | 0.12 16 | 0.20 42 | 0.37 37 | 0.18 26 |
Aniso. Huber-L1 [22] | 44.0 | 0.05 26 | 0.11 40 | 0.05 16 | 0.15 62 | 0.38 62 | 0.17 62 | 0.13 44 | 0.20 51 | 0.17 59 | 0.06 40 | 0.33 59 | 0.06 44 | 0.21 39 | 0.36 38 | 0.16 39 | 0.09 39 | 0.22 30 | 0.09 39 | 0.11 69 | 0.11 51 | 0.18 41 | 0.19 34 | 0.34 32 | 0.20 40 |
Brox et al. [5] | 44.1 | 0.06 41 | 0.15 56 | 0.08 59 | 0.11 48 | 0.33 45 | 0.12 53 | 0.13 44 | 0.21 55 | 0.14 34 | 0.05 15 | 0.28 55 | 0.05 29 | 0.28 62 | 0.44 57 | 0.43 78 | 0.07 10 | 0.27 46 | 0.07 8 | 0.08 27 | 0.10 32 | 0.09 3 | 0.26 64 | 0.63 70 | 0.39 68 |
EP-PM [83] | 44.2 | 0.06 41 | 0.16 63 | 0.06 37 | 0.10 36 | 0.34 49 | 0.09 18 | 0.13 44 | 0.22 58 | 0.13 13 | 0.06 40 | 0.20 43 | 0.07 54 | 0.23 49 | 0.36 38 | 0.16 39 | 0.12 61 | 0.40 66 | 0.12 58 | 0.09 41 | 0.12 62 | 0.37 85 | 0.18 22 | 0.34 32 | 0.17 11 |
LocallyOriented [52] | 45.8 | 0.05 26 | 0.10 34 | 0.05 16 | 0.12 56 | 0.46 68 | 0.12 53 | 0.14 60 | 0.24 61 | 0.16 55 | 0.07 50 | 0.20 43 | 0.06 44 | 0.23 49 | 0.44 57 | 0.17 47 | 0.08 26 | 0.22 30 | 0.09 39 | 0.09 41 | 0.11 51 | 0.17 35 | 0.23 52 | 0.46 53 | 0.26 53 |
CRTflow [88] | 45.9 | 0.06 41 | 0.14 53 | 0.06 37 | 0.10 36 | 0.29 34 | 0.10 28 | 0.12 30 | 0.20 51 | 0.13 13 | 0.06 40 | 0.21 46 | 0.06 44 | 0.20 36 | 0.38 44 | 0.15 33 | 0.24 86 | 0.50 73 | 0.27 84 | 0.08 27 | 0.11 51 | 0.17 35 | 0.25 58 | 0.54 60 | 0.32 62 |
SIOF [69] | 46.0 | 0.07 57 | 0.11 40 | 0.07 50 | 0.10 36 | 0.28 28 | 0.10 28 | 0.15 66 | 0.17 42 | 0.20 66 | 0.08 57 | 0.19 38 | 0.09 62 | 0.24 55 | 0.41 52 | 0.21 55 | 0.11 52 | 0.22 30 | 0.11 50 | 0.09 41 | 0.09 12 | 0.14 23 | 0.25 58 | 0.43 48 | 0.28 57 |
DPOF [18] | 46.7 | 0.06 41 | 0.16 63 | 0.05 16 | 0.11 48 | 0.37 59 | 0.10 28 | 0.13 44 | 0.20 51 | 0.15 51 | 0.07 50 | 0.28 55 | 0.06 44 | 0.21 39 | 0.41 52 | 0.15 33 | 0.09 39 | 0.25 38 | 0.09 39 | 0.09 41 | 0.10 32 | 0.48 88 | 0.25 58 | 0.45 52 | 0.29 59 |
FastOF [78] | 47.3 | 0.06 41 | 0.13 49 | 0.07 50 | 0.10 36 | 0.38 62 | 0.11 46 | 0.13 44 | 0.22 58 | 0.16 55 | 0.05 15 | 0.22 49 | 0.05 29 | 0.23 49 | 0.39 46 | 0.24 63 | 0.12 61 | 0.55 77 | 0.12 58 | 0.09 41 | 0.09 12 | 0.16 29 | 0.23 52 | 0.55 61 | 0.26 53 |
Local-TV-L1 [65] | 47.5 | 0.08 66 | 0.15 56 | 0.11 69 | 0.19 65 | 0.39 64 | 0.22 65 | 0.14 60 | 0.25 64 | 0.17 59 | 0.08 57 | 0.39 65 | 0.09 62 | 0.18 29 | 0.33 32 | 0.13 21 | 0.07 10 | 0.21 27 | 0.07 8 | 0.06 4 | 0.08 2 | 0.18 41 | 0.28 69 | 0.63 70 | 0.49 74 |
NL-TV-NCC [25] | 47.7 | 0.06 41 | 0.11 40 | 0.06 37 | 0.11 48 | 0.34 49 | 0.10 28 | 0.12 30 | 0.15 34 | 0.13 13 | 0.07 50 | 0.22 49 | 0.06 44 | 0.27 61 | 0.45 60 | 0.15 33 | 0.15 72 | 0.36 60 | 0.13 66 | 0.10 60 | 0.14 76 | 0.22 56 | 0.22 49 | 0.41 44 | 0.22 44 |
Dynamic MRF [7] | 48.6 | 0.07 57 | 0.16 63 | 0.08 59 | 0.09 17 | 0.28 28 | 0.09 18 | 0.12 30 | 0.19 47 | 0.14 34 | 0.06 40 | 0.26 53 | 0.06 44 | 0.28 62 | 0.53 71 | 0.22 58 | 0.13 66 | 0.56 78 | 0.14 67 | 0.08 27 | 0.09 12 | 0.24 70 | 0.22 49 | 0.55 61 | 0.27 55 |
CBF [12] | 48.7 | 0.06 41 | 0.13 49 | 0.07 50 | 0.19 65 | 0.36 56 | 0.25 68 | 0.12 30 | 0.17 42 | 0.14 34 | 0.05 15 | 0.24 51 | 0.05 29 | 0.23 49 | 0.39 46 | 0.18 49 | 0.10 45 | 0.22 30 | 0.11 50 | 0.14 79 | 0.14 76 | 0.25 73 | 0.22 49 | 0.42 45 | 0.24 48 |
SuperFlow [89] | 48.8 | 0.05 26 | 0.12 45 | 0.05 16 | 0.15 62 | 0.37 59 | 0.18 63 | 0.13 44 | 0.19 47 | 0.19 64 | 0.10 66 | 0.37 64 | 0.11 66 | 0.23 49 | 0.36 38 | 0.41 76 | 0.06 4 | 0.26 44 | 0.07 8 | 0.10 60 | 0.11 51 | 0.15 25 | 0.26 64 | 0.58 65 | 0.35 64 |
CLG-TV [48] | 50.7 | 0.06 41 | 0.12 45 | 0.06 37 | 0.17 64 | 0.36 56 | 0.21 64 | 0.14 60 | 0.20 51 | 0.19 64 | 0.08 57 | 0.47 71 | 0.08 57 | 0.23 49 | 0.40 49 | 0.21 55 | 0.10 45 | 0.27 46 | 0.11 50 | 0.10 60 | 0.11 51 | 0.13 20 | 0.20 42 | 0.39 40 | 0.21 43 |
Fusion [6] | 51.3 | 0.05 26 | 0.18 68 | 0.06 37 | 0.09 17 | 0.29 34 | 0.09 18 | 0.13 44 | 0.18 44 | 0.14 34 | 0.05 15 | 0.28 55 | 0.05 29 | 0.29 66 | 0.44 57 | 0.30 71 | 0.14 68 | 0.38 64 | 0.15 71 | 0.14 79 | 0.15 81 | 0.19 49 | 0.34 76 | 0.51 58 | 0.41 70 |
Bartels [41] | 51.3 | 0.07 57 | 0.10 34 | 0.08 59 | 0.12 56 | 0.24 13 | 0.14 60 | 0.13 44 | 0.14 30 | 0.17 59 | 0.10 66 | 0.20 43 | 0.11 66 | 0.21 39 | 0.39 46 | 0.20 53 | 0.11 52 | 0.26 44 | 0.16 74 | 0.11 69 | 0.13 69 | 0.32 81 | 0.19 34 | 0.39 40 | 0.22 44 |
Learning Flow [11] | 53.8 | 0.06 41 | 0.13 49 | 0.06 37 | 0.14 61 | 0.37 59 | 0.15 61 | 0.13 44 | 0.21 55 | 0.17 59 | 0.07 50 | 0.34 61 | 0.06 44 | 0.28 62 | 0.55 73 | 0.27 66 | 0.11 52 | 0.38 64 | 0.11 50 | 0.10 60 | 0.12 62 | 0.17 35 | 0.21 46 | 0.50 56 | 0.22 44 |
p-harmonic [29] | 54.6 | 0.07 57 | 0.20 71 | 0.08 59 | 0.11 48 | 0.34 49 | 0.11 46 | 0.14 60 | 0.28 67 | 0.16 55 | 0.06 40 | 0.46 70 | 0.06 44 | 0.30 67 | 0.47 63 | 0.29 69 | 0.11 52 | 0.40 66 | 0.11 50 | 0.10 60 | 0.12 62 | 0.17 35 | 0.20 42 | 0.42 45 | 0.19 34 |
SegOF [10] | 60.0 | 0.08 66 | 0.15 56 | 0.06 37 | 0.36 79 | 0.64 77 | 0.43 79 | 0.18 71 | 0.34 71 | 0.35 80 | 0.15 73 | 0.42 66 | 0.13 68 | 0.42 79 | 0.68 79 | 0.72 83 | 0.10 45 | 0.50 73 | 0.12 58 | 0.06 4 | 0.10 32 | 0.09 3 | 0.23 52 | 0.55 61 | 0.24 48 |
Shiralkar [42] | 60.2 | 0.06 41 | 0.26 78 | 0.08 59 | 0.09 17 | 0.40 66 | 0.09 18 | 0.14 60 | 0.37 72 | 0.14 34 | 0.07 50 | 0.42 66 | 0.07 54 | 0.30 67 | 0.62 76 | 0.20 53 | 0.18 81 | 0.83 80 | 0.17 78 | 0.10 60 | 0.12 62 | 0.33 83 | 0.24 57 | 0.80 77 | 0.27 55 |
Ad-TV-NDC [36] | 60.3 | 0.12 79 | 0.20 71 | 0.20 83 | 0.34 78 | 0.57 75 | 0.41 78 | 0.20 76 | 0.38 73 | 0.27 72 | 0.14 71 | 0.48 73 | 0.16 73 | 0.21 39 | 0.36 38 | 0.16 39 | 0.10 45 | 0.23 36 | 0.10 46 | 0.08 27 | 0.09 12 | 0.16 29 | 0.34 76 | 0.80 77 | 0.81 81 |
Modified CLG [34] | 62.2 | 0.10 72 | 0.23 74 | 0.10 68 | 0.29 75 | 0.56 73 | 0.37 76 | 0.20 76 | 0.52 77 | 0.29 77 | 0.19 78 | 0.72 77 | 0.20 76 | 0.28 62 | 0.53 71 | 0.33 72 | 0.07 10 | 0.37 62 | 0.08 25 | 0.07 14 | 0.10 32 | 0.11 13 | 0.33 75 | 0.83 79 | 0.63 78 |
StereoFlow [44] | 64.4 | 0.30 90 | 0.57 90 | 0.36 88 | 1.03 90 | 1.75 90 | 0.92 87 | 0.81 90 | 1.43 90 | 0.51 87 | 1.05 89 | 2.03 89 | 0.92 88 | 0.53 83 | 0.72 80 | 0.44 79 | 0.04 1 | 0.14 4 | 0.04 1 | 0.04 1 | 0.10 32 | 0.06 1 | 0.27 68 | 0.60 66 | 0.32 62 |
BlockOverlap [61] | 64.6 | 0.12 79 | 0.16 63 | 0.12 74 | 0.25 72 | 0.39 64 | 0.29 73 | 0.19 75 | 0.24 61 | 0.25 69 | 0.16 74 | 0.35 62 | 0.17 75 | 0.19 32 | 0.28 21 | 0.18 49 | 0.15 72 | 0.24 37 | 0.14 67 | 0.15 84 | 0.15 81 | 0.37 85 | 0.25 58 | 0.50 56 | 0.39 68 |
IAOF2 [51] | 66.0 | 0.07 57 | 0.14 53 | 0.07 50 | 0.21 67 | 0.43 67 | 0.25 68 | 0.17 67 | 0.24 61 | 0.22 67 | 0.40 83 | 0.75 79 | 0.66 87 | 0.31 72 | 0.47 63 | 0.27 66 | 0.15 72 | 0.34 57 | 0.16 74 | 0.14 79 | 0.13 69 | 0.20 51 | 0.25 58 | 0.52 59 | 0.29 59 |
HBpMotionGpu [43] | 67.0 | 0.09 68 | 0.15 56 | 0.09 67 | 0.32 77 | 0.49 70 | 0.38 77 | 0.17 67 | 0.32 70 | 0.28 75 | 0.14 71 | 0.35 62 | 0.14 70 | 0.25 58 | 0.40 49 | 0.22 58 | 0.14 68 | 0.27 46 | 0.14 67 | 0.15 84 | 0.13 69 | 0.24 70 | 0.29 70 | 0.60 66 | 0.44 72 |
SPSA-learn [13] | 67.2 | 0.10 72 | 0.26 78 | 0.12 74 | 0.24 71 | 0.50 71 | 0.28 72 | 0.18 71 | 0.40 74 | 0.27 72 | 0.12 68 | 0.55 75 | 0.14 70 | 0.30 67 | 0.47 63 | 0.38 74 | 0.12 61 | 0.41 69 | 0.14 67 | 0.09 41 | 0.11 51 | 0.15 25 | 0.34 76 | 0.64 73 | 0.64 79 |
Filter Flow [19] | 68.0 | 0.09 68 | 0.18 68 | 0.08 59 | 0.22 69 | 0.53 72 | 0.24 67 | 0.17 67 | 0.28 67 | 0.24 68 | 0.16 74 | 0.47 71 | 0.16 73 | 0.32 75 | 0.48 67 | 0.38 74 | 0.17 79 | 0.34 57 | 0.16 74 | 0.18 88 | 0.19 88 | 0.22 56 | 0.23 52 | 0.43 48 | 0.25 52 |
Black & Anandan [4] | 68.5 | 0.10 72 | 0.25 76 | 0.15 80 | 0.23 70 | 0.56 73 | 0.26 70 | 0.18 71 | 0.45 76 | 0.26 71 | 0.13 70 | 0.68 76 | 0.15 72 | 0.31 72 | 0.49 68 | 0.34 73 | 0.12 61 | 0.49 72 | 0.12 58 | 0.11 69 | 0.13 69 | 0.10 9 | 0.30 72 | 0.63 70 | 0.48 73 |
GroupFlow [9] | 68.6 | 0.10 72 | 0.25 76 | 0.14 79 | 0.40 81 | 1.06 84 | 0.44 81 | 0.23 80 | 0.80 82 | 0.37 82 | 0.12 68 | 0.45 68 | 0.13 68 | 0.52 82 | 1.05 87 | 0.24 63 | 0.22 84 | 0.89 82 | 0.25 83 | 0.06 4 | 0.09 12 | 0.09 3 | 0.31 74 | 0.93 81 | 0.43 71 |
2D-CLG [1] | 69.2 | 0.12 79 | 0.28 80 | 0.11 69 | 0.44 83 | 0.72 79 | 0.55 83 | 0.30 83 | 0.74 79 | 0.40 83 | 0.48 86 | 1.17 83 | 0.62 86 | 0.38 77 | 0.67 78 | 0.57 80 | 0.09 39 | 0.51 75 | 0.12 58 | 0.06 4 | 0.09 12 | 0.12 16 | 0.44 82 | 1.04 83 | 0.88 84 |
IAOF [50] | 69.9 | 0.09 68 | 0.18 68 | 0.12 74 | 0.27 74 | 0.48 69 | 0.34 75 | 0.18 71 | 0.44 75 | 0.25 69 | 0.18 77 | 0.53 74 | 0.25 79 | 0.30 67 | 0.50 70 | 0.26 65 | 0.14 68 | 0.53 76 | 0.12 58 | 0.12 75 | 0.11 51 | 0.20 51 | 0.30 72 | 0.65 74 | 0.60 77 |
GraphCuts [14] | 70.1 | 0.10 72 | 0.20 71 | 0.11 69 | 0.21 67 | 0.64 77 | 0.22 65 | 0.17 67 | 0.30 69 | 0.27 72 | 0.09 64 | 0.45 68 | 0.08 57 | 0.30 67 | 0.55 73 | 0.29 69 | 0.16 77 | 0.28 50 | 0.16 74 | 0.13 78 | 0.13 69 | 0.30 79 | 0.37 79 | 0.69 75 | 0.56 75 |
SILK [87] | 73.5 | 0.13 84 | 0.31 81 | 0.21 84 | 0.37 80 | 0.82 80 | 0.43 79 | 0.22 78 | 0.77 81 | 0.31 78 | 0.20 79 | 0.73 78 | 0.22 77 | 0.53 83 | 0.86 83 | 0.79 86 | 0.21 82 | 0.97 83 | 0.21 81 | 0.06 4 | 0.10 32 | 0.15 25 | 0.44 82 | 1.03 82 | 0.86 83 |
Nguyen [33] | 75.0 | 0.12 79 | 0.23 74 | 0.13 78 | 0.60 85 | 0.63 76 | 0.87 86 | 0.26 82 | 0.61 78 | 0.36 81 | 0.40 83 | 0.84 80 | 0.51 84 | 0.40 78 | 0.63 77 | 0.59 81 | 0.17 79 | 0.48 71 | 0.20 80 | 0.10 60 | 0.10 32 | 0.16 29 | 0.40 81 | 0.84 80 | 1.04 85 |
Horn & Schunck [3] | 75.8 | 0.11 77 | 0.35 83 | 0.16 81 | 0.26 73 | 0.87 81 | 0.27 71 | 0.22 78 | 0.83 84 | 0.28 75 | 0.17 76 | 0.87 81 | 0.22 77 | 0.48 80 | 0.78 82 | 0.70 82 | 0.15 72 | 1.07 85 | 0.15 71 | 0.12 75 | 0.14 76 | 0.10 9 | 0.47 84 | 1.26 84 | 0.84 82 |
Periodicity [86] | 78.3 | 0.14 85 | 0.32 82 | 0.11 69 | 0.29 75 | 1.16 86 | 0.31 74 | 0.51 89 | 0.74 79 | 0.66 89 | 0.49 87 | 1.53 87 | 0.43 82 | 1.23 90 | 2.67 90 | 0.95 88 | 0.41 89 | 3.18 90 | 0.37 88 | 0.08 27 | 0.14 76 | 0.09 3 | 0.48 85 | 2.09 89 | 0.79 80 |
SLK [47] | 80.0 | 0.11 77 | 0.50 89 | 0.17 82 | 0.77 87 | 1.27 89 | 1.02 89 | 0.30 83 | 1.11 87 | 0.44 85 | 1.08 90 | 1.23 85 | 1.28 90 | 0.77 88 | 1.07 88 | 1.27 89 | 0.21 82 | 1.22 87 | 0.28 85 | 0.07 14 | 0.14 76 | 0.12 16 | 0.80 89 | 1.44 86 | 1.95 88 |
TI-DOFE [24] | 80.1 | 0.22 88 | 0.42 86 | 0.38 89 | 0.80 88 | 1.16 86 | 0.99 88 | 0.46 88 | 1.15 89 | 0.59 88 | 0.72 88 | 1.38 86 | 0.96 89 | 0.53 83 | 0.86 83 | 0.76 85 | 0.16 77 | 1.12 86 | 0.23 82 | 0.09 41 | 0.12 62 | 0.10 9 | 0.75 88 | 1.45 87 | 1.48 87 |
FOLKI [16] | 82.4 | 0.12 79 | 0.43 87 | 0.22 85 | 0.47 84 | 1.13 85 | 0.69 84 | 0.24 81 | 1.11 87 | 0.32 79 | 0.22 80 | 1.12 82 | 0.29 80 | 0.58 86 | 0.97 85 | 0.90 87 | 0.22 84 | 1.29 89 | 0.32 87 | 0.09 41 | 0.16 84 | 0.28 78 | 0.68 87 | 1.55 88 | 2.32 89 |
Adaptive flow [45] | 83.8 | 0.25 89 | 0.35 83 | 0.33 87 | 0.64 86 | 0.92 82 | 0.73 85 | 0.36 86 | 0.82 83 | 0.47 86 | 0.39 82 | 1.17 83 | 0.43 82 | 0.48 80 | 0.74 81 | 0.42 77 | 0.38 88 | 0.86 81 | 0.38 89 | 0.33 90 | 0.26 90 | 1.16 90 | 0.38 80 | 0.75 76 | 0.56 75 |
PGAM+LK [55] | 85.8 | 0.17 86 | 0.48 88 | 0.23 86 | 0.43 82 | 1.18 88 | 0.50 82 | 0.32 85 | 0.94 86 | 0.41 84 | 0.37 81 | 2.27 90 | 0.38 81 | 0.64 87 | 1.03 86 | 0.73 84 | 0.32 87 | 1.28 88 | 0.28 85 | 0.23 89 | 0.23 89 | 0.49 89 | 0.61 86 | 1.42 85 | 1.13 86 |
Pyramid LK [2] | 87.5 | 0.20 87 | 0.36 85 | 0.40 90 | 0.87 89 | 1.02 83 | 1.40 90 | 0.38 87 | 0.91 85 | 0.71 90 | 0.43 85 | 1.73 88 | 0.61 85 | 0.94 89 | 1.69 89 | 1.39 90 | 0.55 90 | 1.01 84 | 0.50 90 | 0.16 86 | 0.18 87 | 0.32 81 | 1.59 90 | 2.85 90 | 4.45 90 |
Method | time* | frames | color | 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. CVPR 2012. | |
[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] Direct ZNCC | 260 | 2 | color | M. Drulea, C. Pantilie, and S. Nedevschi. A direct approach for correlation-based matching in variational optical flow. Submitted to TIP 2012. | |
[67] ADF | 1535 | 2 | color | Anonymous. Optical flow estimation by adaptive data fusion. NIPS 2012 submission 601. | |
[68] ALD-Flow | 61 | 2 | color | M. Stoll, A. Bruhn, and S. Volz. Adaptive integration of feature matches into variational optic flow methods. ACCV 2012. | |
[69] SIOF | 234 | 2 | color | L. Xu, Z. Dai, and J. Jia. Scale invariant optical flow. ECCV 2012. | |
[70] 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. | |
[71] TCOF | 1421 | all | gray | Anonymous. Optical flow estimation with consistent spatio-temporal coherence models. VISAPP 2013 submission 20. | |
[72] LME | 476 | 2 | color | Anonymous. Optical flow estimation using Laplacian mesh energy. CVPR 2013 submission 11. | |
[73] 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. | |
[74] SCR | 257 | 2 | color | Anonymous. Segmentation constrained regularization for optical flow estimation. CVPR 2013 submission 297. | |
[75] 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. | |
[76] PMF | 35 | 2 | color | Anonymous. PatchMatch filter: efficient edge-aware filtering meets randomized search for fast correspondence field estimation. CVPR 2013 submission 573. | |
[77] 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. | |
[78] FastOF | 0.18 | 2 | color | Anonymous. Quasi-realtime variational optical flow computation. CVPR 2013 submission 792. | |
[79] Correlation Flow | 290 | 2 | color | M. Drulea and S. Nedevschi. Motion estimation using the correlation transform. Submitted to TIP 2013. | |
[80] TC/T-Flow | 341 | 5 | color | Anonymous. Joint trilateral filtering for multiframe optical flow. ICIP 2013 submission 2685. | |
[81] ComplexFlow | 673 | 2 | color | Anonymous. Constructing dense correspondence for complex motion. ICCV 2013 submission 353. | |
[82] OFLADF | 1530 | 2 | color | Anonymous. Optical flow via locally adaptive fusion of complementary data costs. ICCV 2013 submission 423. | |
[83] EP-PM | 2.7 | 2 | color | Anonymous. Fast edge-preserving PatchMatch for large displacement optical flow. ICCV 2013 submission 575. | |
[84] Epistemic | 6.5 | 2 | color | Anonymous. Epistemic optical flow. ICCV 2013 submission 804. | |
[85] Deep-Matching | 13 | 2 | color | Anonymous. Large displacement optical flow with deep matching. ICCV 2013 submission 1095. | |
[86] Periodicity | 8000 | 4 | color | Anonymous. A periodicity-based computation of optical flow. BMVC 2013 submission 133. | |
[87] 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. | |
[88] CRTflow | 13 | 3 | color | Anonymous. The complete rank transform: a tool for accurate and morphologically invariant matching of structures. BMVC 2013 submission 488. | |
[89] SuperFlow | 178 | 2 | color | Anonymous. Superpixel based optical flow estimation. ICCV 2013 submission 507. | |
[90] Levin3 | 247 | 2 | color | L. Chen, J. Wang, and Y. Wu. Decomposing and regularizing sparse/non-sparse components for motion field estimation. Submitted to PAMI 2013. |