Submission ID: 00140
CV-DoG-TFeat-kornia-MAGSAC
Processed: 21-02-05. Download link: sid-00140-cv-dog-tfeat-8k-both-magsac.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: Challenge organizers (contact)
- Keypoint: sift
- Descriptor: tfeat (128 float32: 512 bytes)
- Number of features: 8000
- Summary: OpenCV DoG keypoints, followed by the TFeat descriptor. Implementation: OpenCV + kornia library
- Paper: http://www.bmva.org/bmvc/2016/papers/paper119/paper119.pdf
- Website: N/A
- Origin: Baseline
- Flags: is_baseline
Phototourism / Stereo track
mAA at 10 degrees: 0.46677 (±0.00109 over 3 run(s) / ±0.14566 over 9 scenes)
Rank (per category): 102 (of 147)
| Scene | Features | Matches (matcher) |
Matches (filter) |
Matches (final) |
Rep. @ 3 px. | MS @ 3 px. | mAA(5o) | mAA(10o) |
| bm | 7968.3 | 355.5 | 355.5 | 298.8 | 0.527 Rank: 40/147 |
0.936 Rank: 77/147 |
0.16718 (±0.00202) Rank: 97/147 |
0.28025 (±0.00354) Rank: 96/147 |
| fcs | 7873.7 | 426.3 | 426.3 | 368.6 | 0.443 Rank: 112/147 |
0.932 Rank: 65/147 |
0.57280 (±0.00121) Rank: 98/147 |
0.69694 (±0.00135) Rank: 94/147 |
| lms | 7770.9 | 307.5 | 307.5 | 238.4 | 0.468 Rank: 44/147 |
0.680 Rank: 87/147 |
0.48233 (±0.00121) Rank: 98/147 |
0.58807 (±0.00093) Rank: 99/147 |
| lb | 7952.5 | 288.9 | 288.9 | 224.1 | 0.474 Rank: 113/147 |
0.732 Rank: 100/147 |
0.36153 (±0.00202) Rank: 97/147 |
0.45026 (±0.00127) Rank: 97/147 |
| mc | 7823.8 | 384.7 | 384.7 | 328.2 | 0.524 Rank: 90/147 |
0.946 Rank: 89/147 |
0.34755 (±0.00202) Rank: 97/147 |
0.49405 (±0.00320) Rank: 104/147 |
| mr | 7788.2 | 366.8 | 366.8 | 333.7 | 0.457 Rank: 84/147 |
0.952 Rank: 98/147 |
0.27247 (±0.00153) Rank: 103/147 |
0.37402 (±0.00376) Rank: 104/147 |
| psm | 7807.2 | 176.3 | 176.3 | 136.7 | 0.385 Rank: 93/147 |
0.490 Rank: 115/147 |
0.11902 (±0.00086) Rank: 105/147 |
0.21675 (±0.00076) Rank: 106/147 |
| sf | 7893.5 | 498.4 | 498.4 | 448.3 | 0.480 Rank: 93/147 |
0.866 Rank: 90/147 |
0.45499 (±0.00036) Rank: 104/147 |
0.58478 (±0.00067) Rank: 105/147 |
| spc | 7868.9 | 328.0 | 328.0 | 256.2 | 0.488 Rank: 81/147 |
0.837 Rank: 98/147 |
0.36885 (±0.00192) Rank: 101/147 |
0.51583 (±0.00191) Rank: 99/147 |
| avg | 7860.8 | 348.1 | 348.1 | 292.6 | 0.472 Rank: 82/147 |
0.819 Rank: 106/147 |
0.34964 (±0.00086) Rank: 102/147 |
0.46677 (±0.00109) Rank: 102/147 |
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: 0.62608 (±0.00138 over 3 run(s) / ±0.16776 over 9 scenes)
Rank (per category): 119 (of 147)
| Scene | Features | Matches (input) |
RegistrationRatio (%) | Number of Landmarks |
Track Length | ATE | mAA(50) | mAA(100) |
| bm | 7968.3 | 262.14 | 99.76 Rank: 69/147 |
2107.62 Rank: 130/147 |
4.685 Rank: 120/147 |
0.58564 Rank: 108/147 |
0.34614 (±0.00312) Rank: 113/147 |
0.49636 (±0.00411) Rank: 112/147 |
| fcs | 7873.7 | 339.58 | 98.14 Rank: 72/147 |
3568.39 Rank: 119/147 |
4.272 Rank: 107/147 |
0.27142 Rank: 112/147 |
0.70807 (±0.01222) Rank: 93/147 |
0.78356 (±0.01466) Rank: 80/147 |
| lms | 7770.9 | 241.23 | 97.66 Rank: 131/147 |
2600.47 Rank: 123/147 |
4.167 Rank: 128/147 |
0.29166 Rank: 74/147 |
0.78719 (±0.00736) Rank: 102/147 |
0.84885 (±0.00771) Rank: 104/147 |
| lb | 7952.5 | 262.44 | 97.06 Rank: 119/147 |
2505.96 Rank: 123/147 |
4.316 Rank: 117/147 |
0.51417 Rank: 95/147 |
0.55968 (±0.00638) Rank: 102/147 |
0.65673 (±0.00426) Rank: 102/147 |
| mc | 7823.8 | 273.30 | 98.56 Rank: 133/147 |
3077.63 Rank: 116/147 |
4.210 Rank: 126/147 |
0.45608 Rank: 125/147 |
0.44922 (±0.00431) Rank: 120/147 |
0.58910 (±0.00606) Rank: 126/147 |
| mr | 7788.2 | 290.84 | 91.89 Rank: 127/147 |
3204.09 Rank: 128/147 |
3.819 Rank: 127/147 |
0.51693 Rank: 98/147 |
0.37137 (±0.00833) Rank: 119/147 |
0.48183 (±0.01165) Rank: 120/147 |
| psm | 7807.2 | 104.43 | 77.06 Rank: 137/147 |
1686.34 Rank: 138/147 |
2.908 Rank: 118/147 |
0.77514 Rank: 66/147 |
0.21989 (±0.00696) Rank: 122/147 |
0.29212 (±0.00584) Rank: 127/147 |
| sf | 7893.5 | 381.51 | 95.10 Rank: 134/147 |
4443.31 Rank: 126/147 |
4.040 Rank: 131/147 |
0.40508 Rank: 136/147 |
0.65853 (±0.00546) Rank: 124/147 |
0.74441 (±0.00512) Rank: 129/147 |
| spc | 7868.9 | 234.31 | 98.36 Rank: 109/147 |
2953.45 Rank: 120/147 |
3.924 Rank: 131/147 |
0.57047 Rank: 119/147 |
0.64292 (±0.01015) Rank: 103/147 |
0.74179 (±0.00852) Rank: 105/147 |
| avg | 7860.8 | 265.53 | 94.84 Rank: 134/147 |
2905.25 Rank: 127/147 |
4.038 Rank: 128/147 |
0.48740 Rank: 120/147 |
0.52700 (±0.00073) Rank: 115/147 |
0.62608 (±0.00138) Rank: 119/147 |
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 —
