Optical flow evaluation results       Statistics:     Average   SD   R2.5   R5.0   R10.0   A50   A75   A95  
Error type:   angle   end-point   interpolation   normalized interpolation  
R2.5
interpolation
error		 avg. Army
(Hidden texture)
im0   GT   im1 		Mequon
(Hidden texture)
im0   GT   im1 		Urban
(Synthetic)
im0   GT   im1 		Teddy
(Stereo)
im0   GT   im1 		Backyard
(High-speed camera)
im0   GT   im1 		Basketball
(High-speed camera)
im0   GT   im1 		Dumptruck
(High-speed camera)
im0   GT   im1 		Evergreen
(High-speed camera)
im0   GT   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
Fusion [8]3.8 11.0 8 22.5 10 0.96 6 12.4 1 31.4 2 1.41 2 19.2 4 33.3 4 2.09 4 57.9 5 63.8 4 39.9 1 78.4 4 86.9 4 41.4 2 34.9 1 64.3 2 23.8 2 32.0 5 63.4 3 4.75 3 37.6 1 76.7 1 3.10 12
CBF [14]4.6 10.7 3 21.3 1 0.93 3 12.5 2 31.6 4 1.67 11 18.9 2 32.4 2 2.80 10 57.5 1 62.7 1 40.4 5 78.7 7 87.0 6 42.5 10 35.5 3 63.9 1 24.7 8 31.4 1 62.8 1 4.79 6 37.9 2 76.8 2 3.30 19
F-TV-L1 [17]5.5 10.9 4 22.1 7 1.04 13 13.0 10 32.7 7 1.59 9 19.1 3 34.6 8 2.18 5 57.7 2 63.0 2 40.8 10 78.3 2 86.8 1 41.9 6 36.1 7 65.1 4 24.5 5 31.8 2 63.3 2 4.75 3 38.9 6 78.1 6 3.02 7
TV-L1-improved [19]6.3 10.9 4 21.7 3 0.99 10 12.5 2 31.5 3 1.40 1 19.3 5 33.2 3 2.48 7 57.8 3 63.6 3 40.6 7 78.9 12 87.1 7 42.8 12 36.1 7 65.7 8 25.0 12 32.0 5 65.4 7 5.04 17 38.7 5 77.9 5 3.00 4
GraphCuts [16]6.4 10.5 1 21.5 2 0.94 4 13.8 14 34.7 13 2.08 12 19.3 5 32.3 1 2.95 11 58.8 10 65.6 13 40.5 6 78.7 7 87.1 7 41.9 6 35.0 2 65.0 3 23.5 1 31.8 2 64.4 5 4.92 13 38.0 3 77.6 3 3.06 9
Spatially variant [22]6.5 11.1 9 22.1 7 0.97 9 12.5 2 31.3 1 1.48 6 19.8 11 33.8 6 2.02 2 58.1 6 63.9 5 40.1 2 79.0 13 87.1 7 43.0 14 35.9 4 65.3 6 24.6 6 32.8 16 65.3 6 4.67 1 39.0 8 78.3 8 2.95 2
JIF-Reg [21]7.1 11.2 11 22.9 14 0.96 6 12.7 7 32.7 7 1.47 5 19.5 8 35.4 12 1.90 1 57.8 3 64.0 6 40.2 3 78.8 10 86.9 4 42.7 11 35.9 4 65.2 5 24.6 6 32.6 14 65.5 9 4.71 2 39.4 11 78.8 10 2.94 1
Second-order prior [10]7.6 10.9 4 21.8 4 1.05 15 12.6 5 32.2 5 1.41 2 20.0 12 34.2 7 3.25 17 58.4 9 64.0 6 40.8 10 78.3 2 86.8 1 41.4 2 36.0 6 65.7 8 24.4 3 32.5 13 66.0 11 5.02 16 38.9 6 78.1 6 3.10 12
Brox et al. [7]8.3 11.1 9 22.6 11 0.96 6 12.6 5 32.2 5 1.54 7 19.3 5 34.7 11 2.06 3 58.2 8 64.2 8 40.7 9 79.0 13 87.1 7 43.1 18 36.1 7 65.9 12 24.9 9 32.2 7 65.4 7 4.77 5 39.3 10 78.8 10 3.05 8
DPOF [20]9.6 10.5 1 22.0 6 1.04 13 13.1 11 35.2 14 1.61 10 18.8 1 35.4 12 2.40 6 58.1 6 64.9 10 40.9 13 79.2 18 87.4 17 43.0 14 36.7 18 65.6 7 25.2 16 32.2 7 64.1 4 4.86 10 38.5 4 77.8 4 3.06 9
Dynamic MRF [9]10.1 10.9 4 21.9 5 1.01 11 12.8 8 33.1 10 1.46 4 19.7 10 36.5 17 2.64 8 59.1 13 65.3 12 41.2 15 78.6 5 87.1 7 41.6 4 36.4 10 66.8 19 24.4 3 32.3 9 66.2 13 4.96 14 39.8 14 79.2 13 3.16 14
SPSA-learn [15]11.7 11.5 15 23.3 15 0.90 1 13.7 12 34.0 11 2.15 13 20.5 13 33.7 5 2.99 12 59.0 12 64.7 9 40.2 3 79.0 13 87.2 13 43.0 14 36.4 10 66.2 14 24.9 9 33.3 20 70.4 22 6.17 22 39.2 9 78.7 9 3.00 4
Learning Flow [13]13.9 11.2 11 22.8 12 0.94 4 12.9 9 33.0 9 1.57 8 20.9 15 36.3 16 3.05 14 59.7 15 65.2 11 42.1 18 79.3 21 87.4 17 43.5 21 36.4 10 65.7 8 25.3 17 33.0 17 66.9 16 4.98 15 40.0 17 79.4 15 3.23 17
2D-CLG [3]14.0 12.7 21 25.8 20 1.22 19 13.8 14 34.6 12 2.37 17 19.6 9 34.6 8 2.69 9 59.6 14 66.2 14 40.8 10 79.2 18 87.5 21 43.1 18 36.5 13 66.4 16 25.0 12 32.3 9 66.0 11 4.81 8 39.9 16 79.4 15 3.07 11
SegOF [12]14.3 11.3 13 22.8 12 1.03 12 13.7 12 35.2 14 2.27 15 20.9 15 35.6 14 3.04 13 58.8 10 66.4 16 40.6 7 79.2 18 87.4 17 43.3 20 36.5 13 67.0 20 25.1 14 32.4 12 67.7 19 5.22 19 39.4 11 78.9 12 3.19 16
Black & Anandan [1]14.8 12.0 19 24.5 18 1.07 17 16.2 19 38.6 19 4.17 19 21.6 19 36.5 17 3.50 18 62.9 20 71.2 21 42.6 19 78.6 5 87.1 7 41.8 5