Submission ID: 00540
ASLV2+OANetV2+DEGENSAC
Processed: 20-05-01. Download link: sid-00540-aslv2-8k-oanetv2-degensac.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.
Metadata
- Authors: Jiahui Zhang, Zixin Luo, Hongkai Chen (contact)
- Keypoint: aslv2
- Descriptor: aslv2 (128 float32: 512 bytes)
- Number of features: 8000
- Summary: ASL detector and descriptor, 8k keypoints, using improved OANet matcher and DEGENSAC post-processing
- Paper: N/A
- Website: N/A
- Origin: Submission
- Flags: is_submission, is_challenge_2020
Phototourism / Stereo track
mAA at 10 degrees: 0.50173 (±0.00000 over 3 run(s) / ±0.14936 over 9 scenes)
Rank (per category): 80 (of 147)
Scene | Features | Matches (matcher) |
Matches (filter) |
Matches (final) |
Rep. @ 3 px. | MS @ 3 px. | mAA(5o) | mAA(10o) |
bm | 4268.6 | — | — | 722.7 | 0.601 Rank: 14/147 |
0.816 Rank: 128/147 |
0.13452 (±0.00000) Rank: 115/147 |
0.21593 (±0.00000) Rank: 118/147 |
fcs | 7552.5 | — | — | 1281.3 | 0.547 Rank: 2/147 |
0.791 Rank: 129/147 |
0.54762 (±0.00000) Rank: 111/147 |
0.67419 (±0.00000) Rank: 111/147 |
lms | 3349.0 | — | — | 649.4 | 0.466 Rank: 118/147 |
0.644 Rank: 121/147 |
0.52716 (±0.00000) Rank: 68/147 |
0.65320 (±0.00000) Rank: 56/147 |
lb | 5091.1 | — | — | 583.5 | 0.488 Rank: 72/147 |
0.641 Rank: 125/147 |
0.40259 (±0.00000) Rank: 115/147 |
0.52459 (±0.00000) Rank: 71/147 |
mc | 6257.4 | — | — | 1259.6 | 0.641 Rank: 5/147 |
0.856 Rank: 127/147 |
0.37720 (±0.00000) Rank: 115/147 |
0.54131 (±0.00000) Rank: 54/147 |
mr | 7250.7 | — | — | 1535.1 | 0.614 Rank: 2/147 |
0.868 Rank: 127/147 |
0.26728 (±0.00000) Rank: 107/147 |
0.36658 (±0.00000) Rank: 107/147 |
psm | 7286.7 | — | — | 1081.6 | 0.505 Rank: 2/147 |
0.679 Rank: 40/147 |
0.19739 (±0.00000) Rank: 8/147 |
0.34082 (±0.00000) Rank: 7/147 |
sf | 7090.0 | — | — | 1476.1 | 0.569 Rank: 5/147 |
0.773 Rank: 124/147 |
0.49612 (±0.00000) Rank: 59/147 |
0.62251 (±0.00000) Rank: 72/147 |
spc | 5699.8 | — | — | 810.4 | 0.523 Rank: 14/147 |
0.757 Rank: 123/147 |
0.41746 (±0.00000) Rank: 72/147 |
0.57643 (±0.00000) Rank: 67/147 |
avg | 5982.9 | — | — | 1044.4 | 0.550 Rank: 11/147 |
0.758 Rank: 124/147 |
0.37415 (±0.00000) Rank: 82/147 |
0.50173 (±0.00000) Rank: 80/147 |
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 / Multiview track
mAA at 10 degrees: 0.75275 (±0.00186 over 3 run(s) / ±0.10865 over 9 scenes)
Rank (per category): 13 (of 147)
Scene | Features | Matches (input) |
RegistrationRatio (%) | Number of Landmarks |
Track Length | ATE | mAA(50) | mAA(100) |
bm | 4268.6 | 733.45 | 100.00 Rank: 1/147 |
2746.29 Rank: 93/147 |
4.925 Rank: 84/147 |
0.56660 Rank: 98/147 |
0.41287 (±0.00174) Rank: 84/147 |
0.55959 (±0.00178) Rank: 86/147 |
fcs | 7552.5 | 1328.55 | 98.53 Rank: 42/147 |
5883.50 Rank: 30/147 |
4.843 Rank: 6/147 |
0.24315 Rank: 27/147 |
0.73196 (±0.01072) Rank: 47/147 |
0.79180 (±0.01044) Rank: 65/147 |
lms | 3349.0 | 689.11 | 99.38 Rank: 29/147 |
2575.27 Rank: 127/147 |
5.051 Rank: 3/147 |
0.28281 Rank: 58/147 |
0.84948 (±0.00946) Rank: 13/147 |
0.90376 (±0.00949) Rank: 9/147 |
lb | 5091.1 | 727.77 | 98.68 Rank: 14/147 |
3199.30 Rank: 71/147 |
4.744 Rank: 50/147 |
0.46655 Rank: 9/147 |
0.72783 (±0.00237) Rank: 22/147 |
0.81866 (±0.00280) Rank: 21/147 |
mc | 6257.4 | 1232.29 | 100.00 Rank: 1/147 |
4294.77 Rank: 71/147 |
5.373 Rank: 2/147 |
0.30887 Rank: 4/147 |
0.58468 (±0.00471) Rank: 7/147 |
0.72780 (±0.00431) Rank: 5/147 |
mr | 7250.7 | 1513.21 | 94.96 Rank: 30/147 |
5721.88 Rank: 34/147 |
5.070 Rank: 2/147 |
0.48526 Rank: 49/147 |
0.46207 (±0.00754) Rank: 36/147 |
0.58501 (±0.00846) Rank: 24/147 |
psm | 7286.7 | 1080.86 | 99.98 Rank: 1/147 |
6671.00 Rank: 7/147 |
4.296 Rank: 2/147 |
0.37530 Rank: 4/147 |
0.65190 (±0.00164) Rank: 1/147 |
0.73380 (±0.00106) Rank: 1/147 |
sf | 7090.0 | 1451.46 | 98.98 Rank: 23/147 |
6170.97 Rank: 65/147 |
5.287 Rank: 2/147 |
0.31178 Rank: 34/147 |
0.74970 (±0.00228) Rank: 15/147 |
0.83244 (±0.00267) Rank: 16/147 |
spc | 5699.8 | 847.26 | 99.89 Rank: 15/147 |
4024.26 Rank: 70/147 |
4.947 Rank: 2/147 |
0.43642 Rank: 21/147 |
0.72466 (±0.00198) Rank: 23/147 |
0.82189 (±0.00263) Rank: 22/147 |
avg | 5982.9 | 1067.11 | 98.93 Rank: 6/147 |
4587.47 Rank: 44/147 |
4.948 Rank: 2/147 |
0.38630 Rank: 16/147 |
0.65502 (±0.00204) Rank: 18/147 |
0.75275 (±0.00186) Rank: 13/147 |
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 25 images (displaying: 10). 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 —









