Submission ID: 1551c58b
ytmss
Processed: 21-06-16. Download link: 1551c58b4434d49c-b98323d91215485c-ytmss.json.json
This page ranks the submission against all others using the same number of keypoints, regardless of descriptor size. Please hover over table headers for descriptions on metrics and full scene names.
- Phototourism dataset: Stereo track / Multiview track
- Prague Parks dataset: Stereo track / Multiview track
- Google Urban dataset: Stereo track / Multiview track
Metadata
- Authors: (contact)
- Keypoint: ytmss
- Descriptor: ytmss (2048 float32: 8192 bytes)
- Number of features: 2048
- Summary: ytmss
- Paper: N/A
- Website: N/A
- Processing date: 21-06-16
Phototourism dataset / Stereo track
mAA at 10 degrees: 0.47634 (±0.00000 over 1 run(s) / ±0.13023 over 9 scenes)
Rank (per category): 49 (of 87)
Scene | Features | Matches (raw) |
Matches (final) |
Rep. @ 3 px. | MS @ 3 px. | mAA(5o) | mAA(10o) |
BM | 2043.4 | — | 177.0 | 0.422 Rank: 69/87 |
0.918 Rank: 25/87 |
0.26385 (±0.00000) Rank: 47/87 |
0.40500 (±0.00000) Rank: 45/87 |
FCS | 2048.0 | — | 209.7 | 0.324 Rank: 83/87 |
0.875 Rank: 16/87 |
0.54241 (±0.00000) Rank: 50/87 |
0.67993 (±0.00000) Rank: 48/87 |
LMS | 2045.6 | — | 417.1 | 0.426 Rank: 35/87 |
0.709 Rank: 9/87 |
0.50362 (±0.00000) Rank: 53/87 |
0.62190 (±0.00000) Rank: 55/87 |
LB | 2048.0 | — | 209.0 | 0.360 Rank: 64/87 |
0.746 Rank: 14/87 |
0.42571 (±0.00000) Rank: 47/87 |
0.55136 (±0.00000) Rank: 44/87 |
MC | 2048.0 | — | 165.0 | 0.366 Rank: 79/87 |
0.861 Rank: 38/87 |
0.24177 (±0.00000) Rank: 47/87 |
0.38321 (±0.00000) Rank: 65/87 |
MR | 2038.2 | — | 212.8 | 0.433 Rank: 24/87 |
0.915 Rank: 18/87 |
0.19904 (±0.00000) Rank: 68/87 |
0.29108 (±0.00000) Rank: 68/87 |
PSM | 2048.0 | — | 140.1 | 0.270 Rank: 86/87 |
0.592 Rank: 46/87 |
0.16407 (±0.00000) Rank: 47/87 |
0.30000 (±0.00000) Rank: 45/87 |
SF | 2048.0 | — | 244.1 | 0.361 Rank: 53/87 |
0.829 Rank: 9/87 |
0.36112 (±0.00000) Rank: 66/87 |
0.52582 (±0.00000) Rank: 60/87 |
SPC | 2048.0 | — | 174.9 | 0.325 Rank: 80/87 |
0.810 Rank: 23/87 |
0.35950 (±0.00000) Rank: 54/87 |
0.52880 (±0.00000) Rank: 52/87 |
Avg | 2046.1 | — | 216.6 | 0.365 Rank: 71/87 |
0.806 Rank: 16/87 |
0.34012 (±0.00000) Rank: 53/87 |
0.47634 (±0.00000) Rank: 49/87 |
We show the inliers that survive the robust estimation loop (i.e. RANSAC), or those supplied with the submission if using custom matches, and use the depth estimates to determine whether they are correct. We draw matches above a 5-pixel error threshold in red, and those below are color-coded by their error, from 0 (green) to 5 pixels (yellow). Matches for which we do not have depth estimates are drawn in blue. Please note that the depth maps are estimates and may contain errors.
— British Museum —
— Florence Cathedral Side —
— Lincoln Memorial Statue —
— London Bridge —
— Milan Cathedral —
— Mount Rushmore —
— Piazza San Marco —
— Sagrada Familia —
— Saint Paul's Cathedral —
Phototourism dataset / Multiview track
mAA at 10 degrees: N/A (±0.00000 over 1 run(s) / ±0.11421 over 9 scenes)
Rank (per category): 44 (of 87)
Scene | Features | Matches (input) |
RegistrationRatio (%) | Number of Landmarks |
Track Length | ATE | mAA(50) | mAA(100) |
BM | 2043.4 | 176.88 | 100.00 Rank: 1/87 |
884.08 Rank: 78/87 |
4.643 Rank: 68/87 |
0.41323 Rank: 45/87 |
0.51689 (±0.00000) Rank: 46/87 |
0.67034 (±0.00000) Rank: 46/87 |
FCS | 2048.0 | 212.45 | 97.40 Rank: 48/87 |
1410.78 Rank: 76/87 |
4.257 Rank: 67/87 |
0.24873 Rank: 33/87 |
0.72754 (±0.00000) Rank: 41/87 |
0.78698 (±0.00000) Rank: 43/87 |
LMS | 2045.6 | 441.85 | 99.07 Rank: 51/87 |
1131.41 Rank: 72/87 |
5.889 Rank: 12/87 |
0.30627 Rank: 52/87 |
0.81499 (±0.00000) Rank: 51/87 |
0.86862 (±0.00000) Rank: 54/87 |
LB | 2048.0 | 251.14 | 97.92 Rank: 50/87 |
1058.35 Rank: 70/87 |
4.964 Rank: 57/87 |
0.49146 Rank: 46/87 |
0.69453 (±0.00000) Rank: 42/87 |
0.79242 (±0.00000) Rank: 43/87 |
MC | 2048.0 | 163.30 | 99.88 Rank: 49/87 |
970.15 Rank: 77/87 |
4.486 Rank: 67/87 |
0.35890 Rank: 35/87 |
0.52100 (±0.00000) Rank: 45/87 |
0.67484 (±0.00000) Rank: 43/87 |
MR | 2038.2 | 210.91 | 92.80 Rank: 59/87 |
941.15 Rank: 77/87 |
4.617 Rank: 57/87 |
0.56276 Rank: 62/87 |
0.34775 (±0.00000) Rank: 62/87 |
0.47100 (±0.00000) Rank: 62/87 |
PSM | 2048.0 | 139.74 | 96.34 Rank: 46/87 |
1299.89 Rank: 72/87 |
3.812 Rank: 54/87 |
0.36945 Rank: 39/87 |
0.60757 (±0.00000) Rank: 45/87 |
0.69714 (±0.00000) Rank: 44/87 |
SF | 2048.0 | 240.33 | 99.40 Rank: 42/87 |
1326.18 Rank: 79/87 |
4.651 Rank: 61/87 |
0.29414 Rank: 42/87 |
0.72922 (±0.00000) Rank: 50/87 |
0.81995 (±0.00000) Rank: 45/87 |
SPC | 2048.0 | 179.67 | 100.00 Rank: 1/87 |
1143.57 Rank: 78/87 |
4.511 Rank: 66/87 |
0.51032 Rank: 60/87 |
0.71143 (±0.00000) Rank: 46/87 |
0.82110 (±0.00000) Rank: 42/87 |
Avg | 2046.1 | 224.03 | 98.09 Rank: 47/87 |
1129.51 Rank: 77/87 |
4.648 Rank: 60/87 |
0.39503 Rank: 45/87 |
0.63010 (±0.00000) Rank: 45/87 |
0.73360 (±0.00000) Rank: 44/87 |
In the multi-view track we reconstruct the scene with Structure-from-Motion (Colmap) with small sets of images. We show the results for one bag of 10 images. Keypoints are drawn in blue if they are part of the model, and in red otherwise.
— British Museum —
— Florence Cathedral Side —
— Lincoln Memorial Statue —
— London Bridge —
— Milan Cathedral —
— Mount Rushmore —
— Piazza San Marco —
— Sagrada Familia —
— Saint Paul's Cathedral —
Prague Parks dataset / Stereo track
mAA at 10 degrees: 0.61631 (±0.00000 over 1 run(s) / ±0.02027 over 3 scenes)
Rank (per category): 44 (of 87)
Scene | Features | Matches (raw) |
Matches (final) |
Rep. @ 3 px. | MS @ 3 px. | mAA(5o) | mAA(10o) |
Lizard | 2030.6 | — | 136.8 | 0.026 Rank: 66/87 |
0.009 Rank: 58/87 |
0.45465 (±0.00000) Rank: 45/87 |
0.60472 (±0.00000) Rank: 42/87 |
Pond | 2048.0 | — | 137.6 | 0.048 Rank: 52/87 |
0.050 Rank: 23/87 |
0.44689 (±0.00000) Rank: 44/87 |
0.59940 (±0.00000) Rank: 42/87 |
Tree | 2048.0 | — | 79.6 | 0.022 Rank: 73/87 |
0.021 Rank: 48/87 |
0.50022 (±0.00000) Rank: 52/87 |
0.64481 (±0.00000) Rank: 49/87 |
Avg | 2042.2 | — | 118.0 | 0.032 Rank: 73/87 |
0.027 Rank: 48/87 |
0.46725 (±0.00000) Rank: 46/87 |
0.61631 (±0.00000) Rank: 44/87 |
We show the inliers that survive the robust estimation loop (i.e. RANSAC), or those supplied with the submission if using custom matches, and use the depth estimates to determine whether they are correct. We draw matches above a 5-pixel error threshold in red, and those below are color-coded by their error, from 0 (green) to 5 pixels (yellow). Matches for which we do not have depth estimates are drawn in blue. Please note that the depth maps are estimates and may contain errors.
— Lizard —
— Pond —
— Tree —
Prague Parks dataset / Multiview track
mAA at 10 degrees: N/A (±0.00000 over 1 run(s) / ±0.12591 over 3 scenes)
Rank (per category): 61 (of 87)
Scene | Features | Matches (input) |
RegistrationRatio (%) | Number of Landmarks |
Track Length | ATE | mAA(50) | mAA(100) |
Lizard | 2030.6 | 149.09 | 82.58 Rank: 59/87 |
468.64 Rank: 68/87 |
3.257 Rank: 58/87 |
14.07974 Rank: 24/87 |
0.48569 (±0.00000) Rank: 61/87 |
0.58022 (±0.00000) Rank: 60/87 |
Pond | 2048.0 | 161.51 | 67.27 Rank: 50/87 |
409.44 Rank: 68/87 |
2.824 Rank: 53/87 |
0.42459 Rank: 31/87 |
0.24464 (±0.00000) Rank: 56/87 |
0.27838 (±0.00000) Rank: 57/87 |
Tree | 2048.0 | 78.64 | 66.51 Rank: 67/87 |
201.89 Rank: 69/87 |
2.655 Rank: 68/87 |
5.40484 Rank: 5/87 |
0.33064 (±0.00000) Rank: 63/87 |
0.37447 (±0.00000) Rank: 63/87 |
Avg | 2042.2 | 129.75 | 72.12 Rank: 65/87 |
359.99 Rank: 69/87 |
2.912 Rank: 67/87 |
6.63639 Rank: 12/87 |
0.35366 (±0.00000) Rank: 61/87 |
0.41102 (±0.00000) Rank: 61/87 |
Google Urban dataset / Stereo track
mAA at 10 degrees: 0.35313 (±0.00000 over 1 run(s) / ±0.13190 over 17 scenes)
Rank (per category): 35 (of 87)
Scene | Features | Matches (raw) |
Matches (final) |
mAA(5o) | mAA(10o) |
Amsterdam (AMS) | 2048.0 | — | 130.1 | 0.17946 (±0.00000) Rank: 39/87 |
0.30045 (±0.00000) Rank: 40/87 |
Bangkok (BGK) | 2048.0 | — | 206.3 | 0.05161 (±0.00000) Rank: 37/87 |
0.15012 (±0.00000) Rank: 37/87 |
Barcelona (BCN) | 2048.0 | — | 192.2 | 0.10696 (±0.00000) Rank: 33/87 |
0.19780 (±0.00000) Rank: 32/87 |
Buenos Aires (BAR) | 2048.0 | — | 170.8 | 0.20457 (±0.00000) Rank: 39/87 |
0.34566 (±0.00000) Rank: 37/87 |
Cambridge (CAM) | 2048.0 | — | 152.8 | 0.26370 (±0.00000) Rank: 39/87 |
0.39932 (±0.00000) Rank: 36/87 |
Cannes (CAN) | 2045.6 | — | 198.8 | 0.37880 (±0.00000) Rank: 34/87 |
0.49493 (±0.00000) Rank: 33/87 |
Chicago (CHI) | 2048.0 | — | 101.3 | 0.17322 (±0.00000) Rank: 32/87 |
0.27561 (±0.00000) Rank: 35/87 |
Helsinki (HEL) | 2048.0 | — | 291.1 | 0.31753 (±0.00000) Rank: 39/87 |
0.51598 (±0.00000) Rank: 39/87 |
Madrid (MAD) | 2048.0 | — | 105.9 | 0.06869 (±0.00000) Rank: 38/87 |
0.15471 (±0.00000) Rank: 40/87 |
Mountain View (MTV) | 2048.0 | — | 154.9 | 0.18711 (±0.00000) Rank: 35/87 |
0.32109 (±0.00000) Rank: 36/87 |
New Orleans (NOR) | 2045.3 | — | 131.9 | 0.25385 (±0.00000) Rank: 33/87 |
0.36095 (±0.00000) Rank: 31/87 |
San Francisco (SF) | 2047.9 | — | 180.6 | 0.17743 (±0.00000) Rank: 35/87 |
0.32310 (±0.00000) Rank: 34/87 |
Singapore (SG) | 2043.3 | — | 156.8 | 0.28924 (±0.00000) Rank: 34/87 |
0.40570 (±0.00000) Rank: 35/87 |
Sydney (SYD) | 2048.0 | — | 172.6 | 0.23144 (±0.00000) Rank: 39/87 |
0.38777 (±0.00000) Rank: 40/87 |
Tokyo (TOK) | 2048.0 | — | 199.0 | 0.56127 (±0.00000) Rank: 35/87 |
0.70686 (±0.00000) Rank: 35/87 |
Toronto (TOR) | 2048.0 | — | 134.2 | 0.16240 (±0.00000) Rank: 37/87 |
0.30480 (±0.00000) Rank: 36/87 |
Zurich (ZRH) | 2047.9 | — | 197.8 | 0.26323 (±0.00000) Rank: 24/87 |
0.35842 (±0.00000) Rank: 26/87 |
Average (Avg) | 2047.4 | — | 169.2 | 0.22768 (±0.00000) Rank: 34/87 |
0.35313 (±0.00000) Rank: 35/87 |
We show the inliers that survive the robust estimation loop (i.e. RANSAC), or those supplied with the submission if using custom matches. As the dataset does not contain depth, we color them based on the symmetric epopilar distance on normalized coordinates: from 0 (green) to 2e-4 (yellow), and in red above it. Note that these occasionally contain false positives.
— Amsterdam —
— Bangkok —
— Barcelona —
— Buenos Aires —
— Cambridge —
— Cannes —
— Chicago —
— Helsinki —
— Madrid —
— Mountain View —
— New Orleans —
— San Francisco —
— Singapore —
— Sydney —
— Tokyo —
— Toronto —
— Zurich —
Google Urban dataset / Multiview track
mAA at 10 degrees: N/A (±0.00000 over 1 run(s) / ±0.15195 over 17 scenes)
Rank (per category): 37 (of 87)
Scene | Features | Matches (input) |
RegistrationRatio (%) | Number of Landmarks |
Track Length | ATE | mAA(50) | mAA(100) |
AMS | 2048.0 | 82.76 | 75.78 Rank: 46/87 |
480.93 Rank: 51/87 |
3.250 Rank: 48/87 |
21.51436 Rank: 11/87 |
0.22134 (±0.00000) Rank: 43/87 |
0.29066 (±0.00000) Rank: 44/87 |
BGK | 2048.0 | 142.24 | 95.33 Rank: 36/87 |
1130.42 Rank: 47/87 |
3.546 Rank: 47/87 |
13.34376 Rank: 26/87 |
0.01858 (±0.00000) Rank: 47/87 |
0.05719 (±0.00000) Rank: 48/87 |
BCN | 2048.0 | 114.20 | 81.11 Rank: 38/87 |
681.98 Rank: 40/87 |
3.545 Rank: 42/87 |
25.58010 Rank: 19/87 |
0.12054 (±0.00000) Rank: 33/87 |
0.19753 (±0.00000) Rank: 32/87 |
BAR | 2048.0 | 97.65 | 73.94 Rank: 42/87 |
525.72 Rank: 45/87 |
3.150 Rank: 43/87 |
12.05892 Rank: 9/87 |
0.08781 (±0.00000) Rank: 43/87 |
0.14801 (±0.00000) Rank: 43/87 |
CAM | 2048.0 | 94.63 | 78.55 Rank: 41/87 |
618.62 Rank: 46/87 |
3.160 Rank: 47/87 |
18.95816 Rank: 8/87 |
0.13257 (±0.00000) Rank: 40/87 |
0.18887 (±0.00000) Rank: 41/87 |
CAN | 2045.6 | 108.19 | 71.71 Rank: 34/87 |
423.24 Rank: 39/87 |
3.282 Rank: 39/87 |
16.08416 Rank: 15/87 |
0.05140 (±0.00000) Rank: 35/87 |
0.08544 (±0.00000) Rank: 36/87 |
CHI | 2048.0 | 89.14 | 82.05 Rank: 40/87 |
254.70 Rank: 45/87 |
3.951 Rank: 39/87 |
47.41150 Rank: 28/87 |
0.06410 (±0.00000) Rank: 35/87 |
0.13014 (±0.00000) Rank: 35/87 |
HEL | 2048.0 | 205.13 | 90.67 Rank: 44/87 |
1181.07 Rank: 48/87 |
3.719 Rank: 45/87 |
13.19528 Rank: 16/87 |
0.35230 (±0.00000) Rank: 14/87 |
0.48289 (±0.00000) Rank: 16/87 |
MAD | 2048.0 | 69.31 | 83.93 Rank: 42/87 |
602.69 Rank: 45/87 |
3.173 Rank: 45/87 |
4.39748 Rank: 10/87 |
0.05459 (±0.00000) Rank: 41/87 |
0.12886 (±0.00000) Rank: 40/87 |
MTV | 2048.0 | 110.92 | 89.95 Rank: 38/87 |
802.27 Rank: 42/87 |
3.654 Rank: 43/87 |
17.04284 Rank: 36/87 |
0.25175 (±0.00000) Rank: 34/87 |
0.37457 (±0.00000) Rank: 35/87 |
NOR | 2045.3 | 86.66 | 76.37 Rank: 38/87 |
531.91 Rank: 41/87 |
3.072 Rank: 43/87 |
8.94176 Rank: 11/87 |
0.17913 (±0.00000) Rank: 35/87 |
0.24571 (±0.00000) Rank: 37/87 |
SF | 2047.9 | 115.01 | 90.09 Rank: 40/87 |
837.73 Rank: 43/87 |
3.496 Rank: 42/87 |
19.56924 Rank: 22/87 |
0.22971 (±0.00000) Rank: 36/87 |
0.37269 (±0.00000) Rank: 35/87 |
SG | 2043.3 | 96.61 | 72.59 Rank: 36/87 |
598.90 Rank: 40/87 |
3.157 Rank: 43/87 |
15.66356 Rank: 14/87 |
0.17433 (±0.00000) Rank: 34/87 |
0.23014 (±0.00000) Rank: 34/87 |
SYD | 2048.0 | 103.87 | 84.61 Rank: 38/87 |
823.15 Rank: 41/87 |
3.256 Rank: 42/87 |
18.82238 Rank: 16/87 |
0.27087 (±0.00000) Rank: 27/87 |
0.35594 (±0.00000) Rank: 29/87 |
TOK | 2048.0 | 124.55 | 90.05 Rank: 39/87 |
1008.43 Rank: 46/87 |
3.602 Rank: 41/87 |
13.93856 Rank: 13/87 |
0.60042 (±0.00000) Rank: 39/87 |
0.67350 (±0.00000) Rank: 39/87 |
TOR | 2048.0 | 83.65 | 83.84 Rank: 39/87 |
715.73 Rank: 41/87 |
3.191 Rank: 33/87 |
15.19211 Rank: 13/87 |
0.14158 (±0.00000) Rank: 37/87 |
0.25297 (±0.00000) Rank: 38/87 |
ZRH | 2047.9 | 119.12 | 82.54 Rank: 37/87 |
934.93 Rank: 43/87 |
3.222 Rank: 44/87 |
19.16488 Rank: 26/87 |
0.11133 (±0.00000) Rank: 22/87 |
0.17253 (±0.00000) Rank: 28/87 |
Avg | 2047.4 | 108.45 | 82.54 Rank: 41/87 |
714.85 Rank: 45/87 |
3.378 Rank: 44/87 |
17.69877 Rank: 13/87 |
0.18014 (±0.00000) Rank: 37/87 |
0.25810 (±0.00000) Rank: 37/87 |
In the multi-view track we reconstruct the scene with Structure-from-Motion (Colmap) with small sets of images. We show the results for one bag of 10 images. Keypoints are drawn in blue if they are part of the model, and in red otherwise.
— Amsterdam —
— Bangkok —
— Barcelona —
— Buenos Aires —
— Cambridge —
— Cannes —
— Chicago —
— Helsinki —
— Madrid —
— Mountain View —
— New Orleans —
— San Francisco —
— Singapore —
— Sydney —
— Tokyo —
— Toronto —
— Zurich —