A Comparative Study of Graph Matching Algorithms in Computer Vision

Benchmark Results for “caltech-small”

This page shows the benchmarks results for the dataset “caltech-small”. The reported values, bounds and accuracies are averaged across all instances of the dataset.

Run time 1s

method avg value avg bound feasible optimal accuracy
dd-ls0 -7414.18 -10658.1 21 / 21 9 / 13 58.46%
dd-ls3 -6841.86 -15159.1 21 / 21 7 / 13 56.60%
dd-ls4 -6332.27 -18033.3 21 / 21 5 / 13 53.96%
fgmd inf -inf 0 / 21 0 / 13
fm-bca -8927.49 -12827.5 21 / 21 8 / 13 62.22%
fm -8905.85 -43925.8 21 / 21 11 / 13 57.91%
fw 0 -inf 21 / 21 0 / 13
ga inf -inf 4 / 21 0 / 13
hbp inf -inf 0 / 21 0 / 13
ipfps -8983.02 -inf 21 / 21 4 / 13 67.27%
ipfpu -8829.1 -inf 21 / 21 2 / 13 62.16%
lsm inf -inf 8 / 21 0 / 13
mp -7966.95 -12191.3 21 / 21 3 / 13 59.35%
mp-fw -8886.35 -12793.7 21 / 21 7 / 13 60.29%
mpm inf -inf 9 / 21 0 / 13
mp-mcf -7882.1 -12379.9 21 / 21 1 / 13 59.65%
pm -6509.9 -inf 21 / 21 0 / 13 50.59%
rrwm inf -inf 17 / 21 1 / 13
sm -3931.81 -inf 21 / 21 0 / 13 35.78%
smac -6195.65 -inf 21 / 21 0 / 13 42.38%

Run time 10s

method avg value avg bound feasible optimal accuracy
dd-ls0 -8250.9 -9502.1 21 / 21 10 / 13 59.78%
dd-ls3 -7862.28 -9903.59 21 / 21 11 / 13 60.47%
dd-ls4 -6918.54 -13406.8 21 / 21 8 / 13 53.39%
fgmd inf -inf 4 / 21 3 / 13
fm-bca -8943.17 -12797 21 / 21 9 / 13 62.22%
fm -9039.57 -43925.8 21 / 21 12 / 13 62.44%
fw 0 -inf 21 / 21 0 / 13
ga inf -inf 10 / 21 1 / 13
hbp inf -inf 1 / 21 1 / 13
ipfps -8983.02 -inf 21 / 21 4 / 13 67.27%
ipfpu -8829.1 -inf 21 / 21 2 / 13 62.16%
lsm 0 -inf 21 / 21 0 / 13
mp -8026.49 -12183 21 / 21 3 / 13 60.32%
mp-fw -9051.63 -12244 21 / 21 8 / 13 64.91%
mpm inf -inf 16 / 21 0 / 13
mp-mcf -8029.69 -11736.7 21 / 21 2 / 13 59.27%
pm -6509.9 -inf 21 / 21 0 / 13 50.59%
rrwm -8847.61 -inf 21 / 21 1 / 13 61.43%
sm -3931.81 -inf 21 / 21 0 / 13 35.78%
smac -6195.65 -inf 21 / 21 0 / 13 42.38%

Run time 100s

method avg value avg bound feasible optimal accuracy
dd-ls0 -8250.9 -9502.01 21 / 21 10 / 13 59.78%
dd-ls3 -8357.29 -9465.72 21 / 21 11 / 13 60.86%
dd-ls4 -7877.55 -9809.85 21 / 21 11 / 13 60.12%
fgmd inf -inf 9 / 21 6 / 13
fm-bca -8952.91 -12796.6 21 / 21 10 / 13 60.62%
fm -9049.67 -43925.8 21 / 21 12 / 13 62.72%
fw 0 -inf 21 / 21 0 / 13
ga inf -inf 19 / 21 1 / 13
hbp inf -inf 7 / 21 2 / 13
ipfps -8983.02 -inf 21 / 21 4 / 13 67.27%
ipfpu -8829.1 -inf 21 / 21 2 / 13 62.16%
lsm 0 -inf 21 / 21 0 / 13
mp -8095.4 -12182.3 21 / 21 4 / 13 61.68%
mp-fw -9056.36 -11217.6 21 / 21 8 / 13 64.91%
mpm inf -inf 20 / 21 0 / 13
mp-mcf -8324.87 -10986.8 21 / 21 3 / 13 57.25%
pm -6509.9 -inf 21 / 21 0 / 13 50.59%
rrwm -8847.61 -inf 21 / 21 1 / 13 61.43%
sm -3931.81 -inf 21 / 21 0 / 13 35.78%
smac -6195.65 -inf 21 / 21 0 / 13 42.38%

Run time 300s

method avg value avg bound feasible optimal accuracy
dd-ls0 -8250.9 -9502.01 21 / 21 10 / 13 59.78%
dd-ls3 -8357.29 -9465.72 21 / 21 11 / 13 60.86%
dd-ls4 -8324.16 -9443.03 21 / 21 11 / 13 61.28%
fgmd inf -inf 12 / 21 9 / 13
fm-bca -8960.04 -12796.6 21 / 21 11 / 13 61.10%
fm -9049.67 -43925.8 21 / 21 12 / 13 62.72%
fw 0 -inf 21 / 21 0 / 13
ga inf -inf 20 / 21 1 / 13
hbp inf -inf 8 / 21 2 / 13
ipfps -8983.02 -inf 21 / 21 4 / 13 67.27%
ipfpu -8829.1 -inf 21 / 21 2 / 13 62.16%
lsm 0 -inf 21 / 21 0 / 13
mp -8095.4 -12182.3 21 / 21 4 / 13 61.68%
mp-fw -9056.94 -10819.4 21 / 21 9 / 13 64.38%
mpm inf -inf 20 / 21 0 / 13
mp-mcf -8434.89 -10731.2 21 / 21 3 / 13 56.89%
pm -6509.9 -inf 21 / 21 0 / 13 50.59%
rrwm -8847.61 -inf 21 / 21 1 / 13 61.43%
sm -3931.81 -inf 21 / 21 0 / 13 35.78%
smac -6195.65 -inf 21 / 21 0 / 13 42.38%

Per Instance Results

Results for individual instances of the dataset are also available: