Evaluation Leaderboard
Click a column to sort. Bars indicate relative performance per metric.
Public benchmark
All submissions
Methods are ranked by AUC (VCRE < 45px) by default. Lower-is-better metrics sort ascending first; higher-is-better metrics sort descending first.
Methods evaluated with a single reference image and a single query frame.
| Method | AUC (VCRE < 45px)↕ | AUC (VCRE < 90px)↕ | Precision (VCRE < 45px)↕ | Precision (VCRE < 90px)↕ | Median Reproj. Error (px)↕ | AUC (Err < 25cm, 5°)↕ | Precision (Err < 25cm, 5°)↕ | Median Trans. Error (m)↕ | Median Rot. Error (°)↕ | Estimates for % of frames↕ |
|---|---|---|---|---|---|---|---|---|---|---|
|
DUNE (ViT-Base) + MASt3R
|
0.840
|
0.943
|
64.4%
|
81.1%
|
47.40
|
0.768
|
55.9%
|
0.390
|
4.600
|
99.6%
|
|
DUNE (ViT-Small) + MASt3R
|
0.834
|
0.939
|
63.8%
|
79.5%
|
49.70
|
0.754
|
53.8%
|
0.410
|
4.600
|
99.9%
|
|
MASt3R (Ess.Mat + D.Scale)
|
0.817
|
0.933
|
63.0%
|
79.3%
|
48.80
|
0.740
|
54.7%
|
0.370
|
2.200
|
99.4%
|
|
interp_metric3d_loftr_3d2d
|
0.681
|
0.796
|
39.9%
|
51.4%
|
125.6
|
0.593
|
31.9%
|
1.750
|
31.20
|
100.0%
|
|
Map-Free Visual Relocalization Enhanced by Instance Knowledge and Depth Knowledge
|
0.656
|
0.849
|
45.3%
|
65.0%
|
80.00
|
0.469
|
28.6%
|
0.830
|
11.70
|
99.0%
|
|
DAv3 & SuperGlue (Ess.Mat. + D.Scale)
|
0.641
|
0.780
|
36.0%
|
49.4%
|
126.0
|
0.551
|
27.5%
|
1.690
|
26.70
|
98.5%
|
|
RoMa w/ MicKey depth maps
|
0.604
|
0.734
|
37.8%
|
52.4%
|
111.9
|
0.546
|
31.4%
|
1.180
|
15.60
|
99.9%
|
|
MicKey trained w/ Overlap Score
|
0.572
|
0.748
|
31.2%
|
49.3%
|
129.5
|
0.325
|
13.3%
|
1.660
|
27.30
|
100.0%
|
|
MicKey
|
0.558
|
0.741
|
30.1%
|
49.2%
|
126.9
|
0.283
|
12.0%
|
1.590
|
26.00
|
100.0%
|
|
SuperGlue w/ MicKey depth maps
|
0.556
|
0.711
|
29.8%
|
43.0%
|
139.9
|
0.490
|
23.5%
|
1.700
|
26.10
|
98.5%
|
|
LoFTR w/ MicKey depth maps
|
0.550
|
0.715
|
27.2%
|
40.7%
|
155.0
|
0.467
|
20.3%
|
1.920
|
33.60
|
100.0%
|
|
Mickey Variant GT_Depth
|
0.548
|
0.720
|
28.0%
|
45.5%
|
142.0
|
0.273
|
10.8%
|
1.840
|
30.80
|
100.0%
|
|
FAR (LoFTR)
|
0.481
|
0.680
|
25.3%
|
44.2%
|
137.1
|
0.392
|
17.7%
|
1.480
|
17.30
|
100.0%
|
|
DPT-KITTI & DUSt3R (Ess.Mat. + D.Scale)
|
0.459
|
0.698
|
28.7%
|
50.4%
|
115.8
|
0.394
|
21.4%
|
0.990
|
7.100
|
99.6%
|
|
FAR (SuperGlue)
|
0.456
|
0.668
|
25.8%
|
44.2%
|
135.5
|
0.351
|
17.1%
|
1.500
|
17.30
|
100.0%
|
|
DPT-KITTI & RoMa (Ess.Mat. + D.Scale) Note: maybe buggy
|
0.437
|
0.670
|
26.4%
|
45.7%
|
128.9
|
0.408
|
22.9%
|
1.230
|
11.20
|
100.0%
|
|
KBR++ & LoFTR (PnP)
|
0.420
|
0.635
|
19.4%
|
34.4%
|
165.1
|
0.295
|
11.1%
|
2.240
|
37.80
|
100.0%
|
|
KBR & LoFTR (PnP)
|
0.420
|
0.631
|
19.5%
|
34.0%
|
167.5
|
0.278
|
10.5%
|
2.330
|
39.50
|
100.0%
|
|
DPT-KITTI & ASpanFormer (Ess.Mat. + D.Scale)
|
0.414
|
0.641
|
20.8%
|
37.0%
|
161.8
|
0.361
|
16.3%
|
1.900
|
29.20
|
100.0%
|
|
DPT-KITTI & LoFTR (PnP)
|
0.397
|
0.618
|
18.2%
|
33.5%
|
166.8
|
0.269
|
9.8%
|
2.310
|
39.40
|
100.0%
|
|
DPT-KITTI & LoFTR (Ess.Mat. + D.Scale)
|
0.387
|
0.615
|
18.8%
|
34.7%
|
167.6
|
0.346
|
15.5%
|
1.990
|
30.60
|
100.0%
|
|
DPT-KITTI & SuperGlue (Ess.Mat. + D.Scale)
|
0.385
|
0.602
|
20.6%
|
36.2%
|
160.3
|
0.346
|
16.8%
|
1.880
|
25.40
|
100.0%
|
|
DPT-KITTI & SuperGlue (PnP)
|
0.378
|
0.599
|
19.6%
|
36.0%
|
157.0
|
0.253
|
10.7%
|
2.110
|
32.60
|
100.0%
|
|
DPT-KITTI & LightGlue w/ DISK (Ess.Mat. + D.Scale)
|
0.355
|
0.527
|
19.5%
|
33.2%
|
138.8
|
0.314
|
15.9%
|
1.440
|
18.50
|
78.0%
|
|
DPT-KITTI & DISK (Ess.Mat. + D.Scale)
|
0.346
|
0.536
|
15.1%
|
26.9%
|
208.2
|
0.264
|
10.2%
|
2.590
|
52.00
|
100.0%
|
|
DPT-KITTI & DeDoDe (v1) (Ess.Mat. + D.Scale)
|
0.325
|
0.531
|
16.9%
|
31.2%
|
167.4
|
0.265
|
12.5%
|
2.020
|
30.30
|
88.0%
|
|
DPT-KITTI & SIFT (Ess.Mat. + D.Scale)
|
0.313
|
0.505
|
13.7%
|
25.0%
|
222.9
|
0.253
|
10.3%
|
2.930
|
61.40
|
100.0%
|
|
DPT-KITTI & SIFT (PnP)
|
0.299
|
0.468
|
14.3%
|
25.1%
|
192.0
|
0.191
|
7.8%
|
3.350
|
63.70
|
100.0%
|
|
DPT-KITTI & SiLK (Ess.Mat. + D.Scale)
|
0.192
|
0.312
|
9.8%
|
18.1%
|
176.4
|
0.157
|
7.3%
|
2.210
|
33.80
|
52.2%
|
|
RPR [R(α, β, γ) + s · t(θ, φ)]
|
0.181
|
0.355
|
18.1%
|
35.5%
|
166.4
|
0.105
|
10.5%
|
1.830
|
23.30
|
100.0%
|
|
RPR [3D-3D]
|
0.181
|
0.387
|
18.1%
|
38.7%
|
148.7
|
0.061
|
6.1%
|
1.700
|
22.90
|
100.0%
|
|
RPR [R(6D) +t]
|
0.156
|
0.402
|
15.6%
|
40.2%
|
147.1
|
0.061
|
6.1%
|
1.680
|
22.50
|
100.0%
|
Methods evaluated using sequences of multiple query frames (9-frame window).
| Method | AUC (VCRE < 45px)↕ | AUC (VCRE < 90px)↕ | Precision (VCRE < 45px)↕ | Precision (VCRE < 90px)↕ | Median Reproj. Error (px)↕ | AUC (Err < 25cm, 5°)↕ | Precision (Err < 25cm, 5°)↕ | Median Trans. Error (m)↕ | Median Rot. Error (°)↕ | Estimates for % of frames↕ |
|---|---|---|---|---|---|---|---|---|---|---|
|
[Baseline] MicKey (using only a single frame)Baseline
|
0.572
|
0.752
|
31.3%
|
49.6%
|
128.9
|
0.326
|
13.6%
|
1.660
|
26.90
|
100.0%
|
|
Map-Free Visual Relocalization Enhanced by Instance Knowledge and Depth Knowledge
|
0.387
|
0.721
|
25.2%
|
53.1%
|
105.8
|
0.190
|
10.8%
|
0.950
|
15.10
|
99.1%
|
Click a column header to sort. Bars indicate relative performance per metric.