Submission ID: 00636
MKDNet-cart
Processed: 20-12-13. Download link: sid-00636-sift8k_mkdnet_cart.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-cart (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: 0.46478 (±0.00044 over 3 run(s) / ±0.14878 over 9 scenes)
Rank (per category): 106 (of 147)
| Scene | Features | Matches (matcher) |
Matches (filter) |
Matches (final) |
Rep. @ 3 px. | MS @ 3 px. | mAA(5o) | mAA(10o) |
| bm | 7969.4 | 440.6 | 440.6 | 286.4 | 0.527 Rank: 30/147 |
0.919 Rank: 106/147 |
0.16486 (±0.00219) Rank: 100/147 |
0.26667 (±0.00193) Rank: 101/147 |
| fcs | 7874.2 | 527.3 | 527.3 | 351.5 | 0.443 Rank: 83/147 |
0.931 Rank: 67/147 |
0.57783 (±0.00151) Rank: 92/147 |
0.70183 (±0.00110) Rank: 90/147 |
| lms | 7770.9 | 374.7 | 374.7 | 187.2 | 0.468 Rank: 53/147 |
0.671 Rank: 98/147 |
0.48419 (±0.00181) Rank: 96/147 |
0.59907 (±0.00138) Rank: 94/147 |
| lb | 7953.6 | 345.2 | 345.2 | 199.6 | 0.474 Rank: 86/147 |
0.737 Rank: 95/147 |
0.35951 (±0.00219) Rank: 100/147 |
0.44681 (±0.00172) Rank: 103/147 |
| mc | 7824.0 | 430.2 | 430.2 | 281.1 | 0.524 Rank: 78/147 |
0.936 Rank: 104/147 |
0.33297 (±0.00219) Rank: 100/147 |
0.47624 (±0.00120) Rank: 111/147 |
| mr | 7797.2 | 402.9 | 402.9 | 285.2 | 0.457 Rank: 120/147 |
0.954 Rank: 91/147 |
0.26227 (±0.00237) Rank: 110/147 |
0.36045 (±0.00234) Rank: 110/147 |
| psm | 7809.6 | 246.5 | 246.5 | 136.5 | 0.385 Rank: 83/147 |
0.507 Rank: 110/147 |
0.12552 (±0.00301) Rank: 96/147 |
0.22563 (±0.00235) Rank: 98/147 |
| sf | 7896.6 | 562.8 | 562.8 | 368.6 | 0.480 Rank: 81/147 |
0.868 Rank: 83/147 |
0.44386 (±0.00420) Rank: 109/147 |
0.58142 (±0.00283) Rank: 107/147 |
| spc | 7869.2 | 434.9 | 434.9 | 245.8 | 0.487 Rank: 94/147 |
0.848 Rank: 88/147 |
0.37639 (±0.00023) Rank: 94/147 |
0.52489 (±0.00065) Rank: 95/147 |
| avg | 7862.7 | 418.3 | 418.3 | 260.2 | 0.472 Rank: 91/147 |
0.819 Rank: 107/147 |
0.34749 (±0.00030) Rank: 107/147 |
0.46478 (±0.00044) Rank: 106/147 |
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: 0.66048 (±0.00254 over 3 run(s) / ±0.15079 over 9 scenes)
Rank (per category): 98 (of 147)
| Scene | Features | Matches (input) |
RegistrationRatio (%) | Number of Landmarks |
Track Length | ATE | mAA(50) | mAA(100) |
| bm | 7969.4 | 432.98 | 99.54 Rank: 112/147 |
3128.72 Rank: 65/147 |
4.575 Rank: 128/147 |
0.61689 Rank: 117/147 |
0.35161 (±0.00831) Rank: 112/147 |
0.49891 (±0.00953) Rank: 111/147 |
| fcs | 7874.2 | 555.14 | 98.50 Rank: 50/147 |
4990.46 Rank: 61/147 |
4.283 Rank: 105/147 |
0.25784 Rank: 76/147 |
0.73370 (±0.00628) Rank: 42/147 |
0.80480 (±0.00585) Rank: 31/147 |
| lms | 7770.9 | 411.29 | 98.28 Rank: 119/147 |
4065.23 Rank: 32/147 |
4.023 Rank: 132/147 |
0.28703 Rank: 63/147 |
0.78986 (±0.00372) Rank: 100/147 |
0.84933 (±0.00514) Rank: 103/147 |
| lb | 7953.6 | 422.85 | 97.73 Rank: 105/147 |
3330.44 Rank: 57/147 |
4.343 Rank: 114/147 |
0.53853 Rank: 116/147 |
0.58385 (±0.00658) Rank: 98/147 |
0.68380 (±0.00734) Rank: 95/147 |
| mc | 7824.0 | 420.92 | 99.24 Rank: 114/147 |
3993.17 Rank: 79/147 |
4.264 Rank: 123/147 |
0.45128 Rank: 121/147 |
0.46191 (±0.00206) Rank: 119/147 |
0.60881 (±0.00228) Rank: 118/147 |
| mr | 7797.2 | 407.88 | 92.60 Rank: 113/147 |
4058.87 Rank: 112/147 |
3.899 Rank: 125/147 |
0.53246 Rank: 108/147 |
0.39694 (±0.00532) Rank: 105/147 |
0.50619 (±0.01007) Rank: 106/147 |
| psm | 7809.6 | 240.02 | 89.13 Rank: 92/147 |
3524.04 Rank: 58/147 |
2.891 Rank: 122/147 |
0.66958 Rank: 47/147 |
0.32497 (±0.00774) Rank: 103/147 |
0.41298 (±0.00751) Rank: 102/147 |
| sf | 7896.6 | 558.40 | 98.18 Rank: 47/147 |
5621.28 Rank: 88/147 |
4.157 Rank: 125/147 |
0.32290 Rank: 60/147 |
0.71301 (±0.00527) Rank: 82/147 |
0.80286 (±0.00478) Rank: 64/147 |
| spc | 7869.2 | 445.84 | 99.41 Rank: 53/147 |
4568.02 Rank: 41/147 |
4.016 Rank: 118/147 |
0.49051 Rank: 80/147 |
0.67020 (±0.00550) Rank: 83/147 |
0.77664 (±0.00539) Rank: 80/147 |
| avg | 7862.7 | 432.81 | 96.96 Rank: 90/147 |
4142.25 Rank: 70/147 |
4.050 Rank: 127/147 |
0.46300 Rank: 104/147 |
0.55845 (±0.00205) Rank: 97/147 |
0.66048 (±0.00254) Rank: 98/147 |
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 —
