Submission ID: 00638
MKDNet-polar
Processed: 20-12-11. Download link: sid-00638-sift8k_mkdnet_polar.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: (contact)
- Keypoint: sift
- Descriptor: mkdnet-polar (128 float32: 512 bytes)
- Number of features: 8000
- Summary: based on the paper [Explicit spatial encoding for deep local descriptors], trained on Liberty set from PhotoTourism dataset
- Paper: https://openaccess.thecvf.com/content_CVPR_2019/papers/Mukundan_Explicit_Spatial_Encoding_for_Deep_Local_Descriptors_CVPR_2019_paper.pdf
- Website: https://openaccess.thecvf.com/content_CVPR_2019/papers/Mukundan_Explicit_Spatial_Encoding_for_Deep_Local_Descriptors_CVPR_2019_paper.pdf
- Origin: Submission
- Flags: is_submission
Phototourism / Stereo track
mAA at 10 degrees: 0.51376 (±0.00106 over 3 run(s) / ±0.15039 over 9 scenes)
Rank (per category): 70 (of 147)
| Scene | Features | Matches (matcher) |
Matches (filter) |
Matches (final) |
Rep. @ 3 px. | MS @ 3 px. | mAA(5o) | mAA(10o) |
| bm | 7969.4 | 572.0 | 572.0 | 376.3 | 0.527 Rank: 30/147 |
0.938 Rank: 72/147 |
0.20306 (±0.00182) Rank: 72/147 |
0.31761 (±0.00275) Rank: 71/147 |
| fcs | 7874.2 | 674.5 | 674.5 | 452.0 | 0.443 Rank: 83/147 |
0.941 Rank: 29/147 |
0.61929 (±0.00133) Rank: 51/147 |
0.74025 (±0.00138) Rank: 49/147 |
| lms | 7770.9 | 471.2 | 471.2 | 244.4 | 0.468 Rank: 53/147 |
0.694 Rank: 61/147 |
0.53452 (±0.00299) Rank: 58/147 |
0.65378 (±0.00196) Rank: 54/147 |
| lb | 7953.6 | 428.0 | 428.0 | 246.5 | 0.474 Rank: 86/147 |
0.766 Rank: 51/147 |
0.42205 (±0.00182) Rank: 72/147 |
0.51290 (±0.00112) Rank: 77/147 |
| mc | 7824.0 | 576.2 | 576.2 | 378.0 | 0.524 Rank: 78/147 |
0.952 Rank: 72/147 |
0.36802 (±0.00182) Rank: 72/147 |
0.51732 (±0.00321) Rank: 87/147 |
| mr | 7797.2 | 518.5 | 518.5 | 370.9 | 0.457 Rank: 120/147 |
0.968 Rank: 60/147 |
0.29939 (±0.00174) Rank: 87/147 |
0.40698 (±0.00101) Rank: 86/147 |
| psm | 7809.6 | 304.6 | 304.6 | 174.8 | 0.385 Rank: 83/147 |
0.596 Rank: 79/147 |
0.14658 (±0.00230) Rank: 71/147 |
0.26002 (±0.00114) Rank: 71/147 |
| sf | 7896.6 | 712.1 | 712.1 | 469.9 | 0.480 Rank: 81/147 |
0.894 Rank: 20/147 |
0.48666 (±0.00175) Rank: 73/147 |
0.62746 (±0.00194) Rank: 62/147 |
| spc | 7869.2 | 560.9 | 560.9 | 324.2 | 0.487 Rank: 94/147 |
0.876 Rank: 45/147 |
0.43625 (±0.00319) Rank: 58/147 |
0.58750 (±0.00330) Rank: 58/147 |
| avg | 7862.7 | 535.3 | 535.3 | 337.4 | 0.472 Rank: 91/147 |
0.847 Rank: 51/147 |
0.39065 (±0.00125) Rank: 70/147 |
0.51376 (±0.00106) Rank: 70/147 |
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: 0.69907 (±0.00048 over 3 run(s) / ±0.13701 over 9 scenes)
Rank (per category): 63 (of 147)
| Scene | Features | Matches (input) |
RegistrationRatio (%) | Number of Landmarks |
Track Length | ATE | mAA(50) | mAA(100) |
| bm | 7969.4 | 565.23 | 99.71 Rank: 85/147 |
3608.96 Rank: 26/147 |
4.810 Rank: 113/147 |
0.53602 Rank: 89/147 |
0.41889 (±0.00573) Rank: 81/147 |
0.57022 (±0.00771) Rank: 83/147 |
| fcs | 7874.2 | 707.24 | 98.33 Rank: 64/147 |
5841.97 Rank: 31/147 |
4.387 Rank: 88/147 |
0.25219 Rank: 58/147 |
0.73298 (±0.01385) Rank: 45/147 |
0.79522 (±0.01631) Rank: 57/147 |
| lms | 7770.9 | 513.62 | 99.36 Rank: 39/147 |
4400.83 Rank: 25/147 |
4.332 Rank: 111/147 |
0.29802 Rank: 88/147 |
0.83293 (±0.01017) Rank: 40/147 |
0.89148 (±0.01087) Rank: 35/147 |
| lb | 7953.6 | 522.51 | 98.65 Rank: 16/147 |
3790.85 Rank: 28/147 |
4.491 Rank: 95/147 |
0.50618 Rank: 85/147 |
0.64133 (±0.00914) Rank: 71/147 |
0.73688 (±0.00963) Rank: 74/147 |
| mc | 7824.0 | 564.82 | 99.81 Rank: 43/147 |
4714.64 Rank: 45/147 |
4.490 Rank: 94/147 |
0.42249 Rank: 103/147 |
0.48945 (±0.00706) Rank: 94/147 |
0.63769 (±0.00835) Rank: 97/147 |
| mr | 7797.2 | 524.34 | 94.25 Rank: 72/147 |
4915.70 Rank: 63/147 |
4.101 Rank: 79/147 |
0.50779 Rank: 87/147 |
0.44166 (±0.00523) Rank: 56/147 |
0.55489 (±0.00786) Rank: 56/147 |
| psm | 7809.6 | 297.08 | 93.02 Rank: 60/147 |
4176.05 Rank: 38/147 |
3.075 Rank: 104/147 |
0.60292 Rank: 33/147 |
0.37436 (±0.00185) Rank: 80/147 |
0.47081 (±0.00311) Rank: 66/147 |
| sf | 7896.6 | 704.39 | 98.97 Rank: 24/147 |
6470.03 Rank: 55/147 |
4.370 Rank: 77/147 |
0.30958 Rank: 32/147 |
0.73887 (±0.00530) Rank: 31/147 |
0.82295 (±0.00502) Rank: 29/147 |
| spc | 7869.2 | 576.93 | 99.73 Rank: 29/147 |
5161.29 Rank: 25/147 |
4.276 Rank: 77/147 |
0.46866 Rank: 56/147 |
0.71110 (±0.00288) Rank: 39/147 |
0.81151 (±0.00360) Rank: 30/147 |
| avg | 7862.7 | 552.91 | 97.98 Rank: 43/147 |
4786.70 Rank: 29/147 |
4.259 Rank: 96/147 |
0.43376 Rank: 71/147 |
0.59795 (±0.00086) Rank: 65/147 |
0.69907 (±0.00048) Rank: 63/147 |
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 —
