Benchmark

Listed below are the public results on the three benchmark datasets. The localization results are reported as the percentage of query images which where localized within three given translation and rotation thresholds, for each condition.

Aachen Day-Night dataset

Localization thresholds:

  • Day: (0.25m, 2°) / (0.5m, 5°) / (5m, 10°)
  • Night: (0.5m, 2°) / (1m, 5°) / (5m, 10°)
MethoddaynightLast updated
ONavi85.7 / 93.7 / 98.948.0 / 71.4 / 88.8May 21, 2020, 2:39 a.m.
Hierarchical-Localization + SuperGlue89.6 / 96.1 / 98.844.9 / 71.4 / 88.8May 25, 2020, 6:04 a.m.
Visual Localization Using Dense Semantic 3D Map And Hybrid Features90.3 / 95.5 / 97.944.9 / 67.3 / 87.8May 25, 2020, 6:26 a.m.
KAPTURE-R2D2-APGeM88.7 / 95.8 / 98.844.9 / 62.2 / 85.7May 24, 2020, 8:42 p.m.
rkpd2m_5k87.7 / 93.7 / 97.042.9 / 66.3 / 85.7April 22, 2020, 8:45 a.m.
NetVLAD (top-20) & D2-Net - multi-scale84.8 / 92.6 / 97.543.9 / 66.3 / 85.7Nov. 1, 2019, 9:49 a.m.
spd2ms88.2 / 93.2 / 97.042.9 / 66.3 / 84.7April 22, 2020, 8:44 a.m.
Geometric Prior Guided Camera Localization86.0 / 92.8 / 97.642.9 / 63.3 / 85.7March 1, 2020, 10:15 a.m.
RT_AP_IR+CRBNet76.2 / 93.0 / 98.941.8 / 64.3 / 85.7March 6, 2020, 2:37 a.m.
D2-net84.3 / 91.9 / 96.243.9 / 68.4 / 84.7May 22, 2020, 8:54 a.m.
ONavi-H77.7 / 89.0 / 95.849.0 / 65.3 / 85.7Sept. 24, 2019, 9:31 a.m.
nv-sp-gp83.9 / 92.0 / 97.646.9 / 63.3 / 84.7May 25, 2020, 6:51 a.m.
GO_r2d2_best84.2 / 92.6 / 97.042.9 / 63.3 / 85.7May 24, 2020, 4:15 p.m.
GO_R2D282.9 / 90.2 / 95.842.9 / 63.3 / 85.7May 20, 2020, 2:05 a.m.
UR2KID (global: ur2kid-top20, local: ur2kid-single-scale)79.9 / 88.6 / 93.645.9 / 64.3 / 83.7Nov. 29, 2019, 4:36 p.m.
IR30+ML77.1 / 90.5 / 97.543.9 / 61.2 / 82.7Aug. 2, 2019, 12:31 p.m.
UP_RD2_NV84.7 / 92.1 / 96.642.9 / 59.2 / 77.6April 16, 2020, 9:56 a.m.
Visual Localization Using Sparse Semantic 3D Map71.8 / 91.5 / 96.840.8 / 63.3 / 80.6June 2, 2019, 4 a.m.
DenseVLAD (top-50) & D2-Net - multi-scale83.1 / 90.9 / 95.540.8 / 62.2 / 80.6Nov. 1, 2019, 9:49 a.m.
ONavi0.0 / 0.0 / 0.048.0 / 71.4 / 88.8May 21, 2020, 2:40 a.m.
Fine_NV_SP81.1 / 89.4 / 95.540.8 / 60.2 / 80.6April 6, 2020, 2:56 p.m.
Hierarchical-Localization NetVLAD+SuperPoint80.5 / 87.4 / 94.242.9 / 62.2 / 76.5June 1, 2019, 9:50 a.m.
Hierarchical-Localization (multi-camera when available)80.5 / 87.4 / 94.242.9 / 62.2 / 76.5June 2, 2019, 6:55 a.m.
ALL_NV_SP79.6 / 87.1 / 93.839.8 / 61.2 / 77.6April 3, 2020, 10:42 a.m.
Fine_GeM_SP78.5 / 87.3 / 93.841.8 / 58.2 / 75.5April 9, 2020, 11:41 a.m.
ASLFeat (10k) + OANet0.0 / 0.0 / 0.048.0 / 67.3 / 88.8March 20, 2020, 4:57 a.m.
R2D2_40K0.0 / 0.0 / 0.048.0 / 67.3 / 88.8May 18, 2020, 4:15 p.m.
DenseVLAD & D2-Net (top-20)80.1 / 88.0 / 93.439.8 / 55.1 / 74.5Oct. 31, 2019, 12:58 a.m.
SuperPoint + SuperGlue0.0 / 0.0 / 0.045.9 / 70.4 / 88.8March 18, 2020, 3:14 p.m.
sd2ss_zoo4cro_supersc_p20.0 / 0.0 / 0.045.9 / 70.4 / 88.8May 24, 2020, 7:24 p.m.
UR2KID Scape Technologies (paper coming soon, not scene specific, single scale)0.0 / 0.0 / 0.046.9 / 67.3 / 88.8July 3, 2019, 3:39 p.m.
PoseDesc+SuperPoint kp0.0 / 0.0 / 0.045.9 / 69.4 / 88.8March 13, 2020, 5:34 p.m.
d2net5190.0 / 0.0 / 0.045.9 / 69.4 / 88.8May 19, 2020, 12:32 p.m.
D2-Net - single-scale0.0 / 0.0 / 0.045.9 / 68.4 / 88.8July 16, 2019, 12:21 p.m.
R2D2 V2 20K0.0 / 0.0 / 0.046.9 / 66.3 / 88.8May 18, 2020, 4:14 p.m.
CA-R2D20.0 / 0.0 / 0.048.0 / 65.3 / 88.8May 22, 2020, 11:04 a.m.
sd2ss_superSC_all0.0 / 0.0 / 0.044.9 / 71.4 / 88.8May 23, 2020, 11:56 a.m.
coarse2fine73.2 / 84.5 / 92.831.6 / 54.1 / 76.5May 24, 2020, 2:03 p.m.
R2D2 10k keypoints0.0 / 0.0 / 0.045.9 / 66.3 / 88.8June 17, 2019, 9:56 a.m.
Active Search v1.185.3 / 92.2 / 97.927.6 / 38.8 / 56.1April 26, 2020, 9:27 a.m.
Dense-ContextDesc10k_upright_OANet0.0 / 0.0 / 0.048.0 / 63.3 / 88.8Feb. 27, 2020, 7:33 a.m.
sgnet-vl0.0 / 0.0 / 0.046.9 / 65.3 / 88.8Oct. 1, 2019, 11:47 a.m.
HSC-Net71.1 / 81.9 / 91.732.7 / 43.9 / 65.3Nov. 27, 2019, 7:09 p.m.
Cascaded Parallel Filtering for Memory-Efficient Image-based Localization76.7 / 88.6 / 95.825.5 / 38.8 / 54.1July 19, 2019, 10:10 a.m.
Asymmetric Hypercolumn Matching47.8 / 72.2 / 91.330.6 / 53.1 / 78.6June 5, 2019, 7:28 p.m.
DFC-Net0.0 / 0.0 / 0.044.9 / 67.3 / 88.8Sept. 19, 2019, 12:56 p.m.
hpf0.0 / 0.0 / 0.044.9 / 70.4 / 87.8May 25, 2020, 8:02 a.m.
densecontextdesc10k_upright_mixedmatcher0.0 / 0.0 / 0.046.9 / 65.3 / 87.8July 16, 2019, 11:13 a.m.
PoseDesc+SIFT kp0.0 / 0.0 / 0.044.9 / 68.4 / 87.8March 13, 2020, 5:35 p.m.
swnet-single0.0 / 0.0 / 0.043.9 / 64.3 / 87.8Oct. 17, 2019, 9 a.m.
D2-Net - single-scale off-the-shelf0.0 / 0.0 / 0.041.8 / 69.4 / 86.7July 16, 2019, 4:57 p.m.
NetVLAD+SuperPoint top 10 (baseline)33.9 / 48.8 / 78.525.5 / 45.9 / 80.6June 1, 2019, 8:59 p.m.
r2d25190.0 / 0.0 / 0.044.9 / 64.3 / 85.7May 19, 2020, 8:48 a.m.
SOSNet+D2D0.0 / 0.0 / 0.042.9 / 64.3 / 85.7Dec. 10, 2019, 4:34 p.m.
nv-sp-gp0.0 / 0.0 / 0.046.9 / 63.3 / 84.7May 25, 2020, 7:01 a.m.
CityScaleLocalization52.3 / 80.0 / 94.324.5 / 33.7 / 49.0May 10, 2019, 2:22 p.m.
Localizing Visual Landmarks for 2D matching62.4 / 71.8 / 79.924.5 / 35.7 / 44.9April 20, 2019, 5:21 p.m.
DELF - new model0.0 / 0.0 / 0.039.8 / 61.2 / 85.7June 12, 2019, 3:52 p.m.
HardNet+D2D0.0 / 0.0 / 0.041.8 / 61.2 / 84.7Dec. 10, 2019, 4:32 p.m.
Active Search57.3 / 83.7 / 96.619.4 / 30.6 / 43.9March 13, 2019, 9 a.m.
DELF - old model0.0 / 0.0 / 0.039.8 / 60.2 / 84.7May 21, 2019, 2:52 p.m.
saliency_ranking_net (multi-scale)0.0 / 0.0 / 0.044.9 / 59.2 / 77.6April 29, 2019, 2:33 p.m.
DGCNCCC22.9 / 49.8 / 84.719.4 / 37.8 / 68.4May 30, 2019, 10:17 a.m.
densecontextdesc10k_upright0.0 / 0.0 / 0.041.8 / 57.1 / 79.6Aug. 27, 2019, 11:11 a.m.
superpoint5190.0 / 0.0 / 0.040.8 / 59.2 / 78.6May 19, 2020, 8:44 a.m.
SuperPoint (baseline)0.0 / 0.0 / 0.042.9 / 57.1 / 77.6July 18, 2019, 8:24 a.m.
contextdesc10_upright0.0 / 0.0 / 0.040.8 / 55.1 / 80.6Aug. 27, 2019, 11:10 a.m.
HesAffNet-HardNet20.0 / 0.0 / 0.037.8 / 54.1 / 75.5May 31, 2019, 10:35 p.m.
ESAC (50 Experts)42.6 / 59.6 / 75.53.1 / 9.2 / 11.2Aug. 8, 2019, 7:14 a.m.
Upright RootSIFT (Feature Challenge Baseline)0.0 / 0.0 / 0.033.7 / 52.0 / 65.3May 2, 2019, 8:57 a.m.
ESAC (20 Experts)39.7 / 55.9 / 77.82.0 / 8.2 / 11.2Aug. 8, 2019, 7:13 a.m.
ESAC (10 Experts)30.3 / 49.3 / 73.72.0 / 5.1 / 7.1Aug. 8, 2019, 7:13 a.m.
Superpoint+PQ46.2 / 65.9 / 86.50.0 / 0.0 / 0.0Sept. 17, 2019, 9:30 a.m.
m20.0 / 0.0 / 0.027.6 / 42.9 / 60.2Nov. 14, 2019, 11:38 p.m.
ELF0.0 / 0.0 / 0.013.3 / 21.4 / 30.6May 3, 2019, 8:36 p.m.
DSAC++0.4 / 2.4 / 34.00.0 / 0.0 / 3.1Aug. 8, 2019, 7:15 a.m.
DenseVLAD0.0 / 0.1 / 22.80.0 / 2.0 / 14.3April 11, 2019, 5:43 p.m.
NetVLAD0.0 / 0.2 / 18.90.0 / 2.0 / 12.2May 31, 2019, 11:16 p.m.
Localizing Visual Landmarks for Place Recognition0.0 / 0.2 / 20.80.0 / 1.0 / 10.2June 2, 2019, 10:06 p.m.
NetVlad_NN_Pose0.0 / 0.1 / 7.00.0 / 0.0 / 4.1March 17, 2019, 11:46 a.m.
joint image segemation and depth to posenet0.0 / 0.1 / 24.90.0 / 0.0 / 0.0May 31, 2019, 11:45 a.m.
FAB-MAP0.0 / 0.0 / 4.60.0 / 0.0 / 0.0March 13, 2019, 11:54 a.m.

CMU Seasons dataset

Localization thresholds:

  • All conditions: (0.25m, 2°) / (0.5m, 5°) / (5m, 10°)
MethodurbansuburbanparkLast updated
Geometric Prior Guided Camera Localization93.5 / 96.0 / 98.381.1 / 87.6 / 94.567.1 / 76.1 / 86.1March 1, 2020, 10:14 a.m.
UniGAN(RGBS)+NV+SP92.4 / 95.0 / 98.075.9 / 82.1 / 91.056.8 / 65.1 / 81.7Jan. 20, 2020, 7:25 a.m.
UniGAN(RGB)+NV+SP92.1 / 94.8 / 98.075.9 / 81.6 / 90.855.5 / 64.1 / 81.0Jan. 20, 2020, 7:26 a.m.
Hierarchical-Localization NetVLAD+SuperPoint91.7 / 94.6 / 97.774.5 / 81.5 / 91.354.3 / 62.5 / 79.0June 1, 2019, 9:50 a.m.
LocalSfM (Upright RootSIFT) (has partial access to ground truth)72.8 / 74.1 / 76.155.2 / 57.7 / 61.341.8 / 44.5 / 48.7March 13, 2019, 11:42 a.m.
Semantic Match Consistency75.0 / 82.1 / 87.844.0 / 53.6 / 63.730.0 / 37.9 / 48.2March 18, 2019, 11:47 p.m.
Active Search v1.168.9 / 75.7 / 83.436.2 / 44.4 / 56.024.8 / 31.1 / 41.5Aug. 12, 2019, 6:38 a.m.
EffecientPairWise2D-3DMatching64.1 / 67.8 / 71.530.2 / 35.5 / 40.923.8 / 28.1 / 34.0March 20, 2019, 9:52 a.m.
coarse2fine22.7 / 47.6 / 88.811.0 / 27.3 / 76.512.5 / 31.0 / 78.4April 1, 2020, 3:42 p.m.
DenseVLAD22.2 / 48.6 / 92.89.8 / 26.6 / 85.210.3 / 27.1 / 77.0April 15, 2019, 11:03 a.m.
Active Search55.2 / 60.3 / 65.120.7 / 25.9 / 29.912.7 / 16.3 / 20.8March 13, 2019, 9:04 a.m.
DIFL+FCL (urban model and square-resized)20.2 / 44.7 / 87.89.7 / 25.0 / 73.711.4 / 28.9 / 76.4Sept. 25, 2019, 11:14 a.m.
DIFL+FCL20.2 / 45.0 / 87.19.1 / 23.3 / 69.310.1 / 26.4 / 74.0Sept. 25, 2019, 11:03 a.m.
NetVLAD17.4 / 40.3 / 93.27.6 / 21.0 / 80.55.6 / 15.7 / 65.8April 11, 2019, 9:16 p.m.
CityScaleLocalization36.7 / 42.0 / 53.18.6 / 11.7 / 21.17.0 / 9.6 / 17.0March 13, 2019, 11:59 a.m.
FAB-MAP2.7 / 6.4 / 27.30.5 / 1.5 / 13.60.8 / 1.7 / 11.5March 13, 2019, 11:49 a.m.

Extended CMU Seasons dataset

Localization thresholds:

  • All conditions: (0.25m, 2°) / (0.5m, 5°) / (5m, 10°)
MethodurbansuburbanparkLast updated
PFSL FGSN 200 classes (sequential localization)94.1 / 99.3 / 100.0100.0 / 100.0 / 100.097.6 / 99.9 / 99.9Nov. 4, 2019, 1:37 p.m.
PFSL Cityscapes classes (sequential localization)88.1 / 93.8 / 100.097.4 / 99.2 / 100.081.8 / 97.2 / 99.9Nov. 4, 2019, 1:40 p.m.
DenseVLAD & D2-Net94.0 / 97.7 / 99.193.0 / 95.7 / 98.389.2 / 93.2 / 95.0Feb. 25, 2020, 9:10 p.m.
KAPTURE-R2D2-APGeM96.7 / 98.9 / 99.794.4 / 96.8 / 99.283.6 / 89.0 / 95.5May 24, 2020, 3:10 p.m.
Hierarchical-Localization (multi-camera when available)91.6 / 96.4 / 99.184.7 / 91.5 / 98.669.3 / 77.8 / 90.5June 2, 2019, 6:55 a.m.
Visual Localization Using Sparse Semantic 3D Map88.8 / 93.6 / 96.378.0 / 83.8 / 89.263.6 / 70.3 / 77.3June 2, 2019, 4 a.m.
CU-LTVL90.3 / 94.8 / 97.876.9 / 83.0 / 91.857.5 / 64.3 / 78.8Feb. 20, 2020, 9:05 a.m.
Hierarchical-Localization NetVLAD+SuperPoint89.5 / 94.2 / 97.976.5 / 82.7 / 92.757.4 / 64.4 / 80.4June 1, 2019, 9:50 a.m.
Asymmetric Hypercolumn Matching65.7 / 82.7 / 91.066.5 / 82.6 / 92.954.3 / 71.6 / 84.1June 5, 2019, 7:24 p.m.
S2DHM feature optimization62.9 / 81.9 / 95.955.1 / 77.6 / 94.750.1 / 72.0 / 88.1May 24, 2020, 12:43 p.m.
Match Consistency with Fine-Grained Segmentations (FGSN+SSMC)85.3 / 91.0 / 94.669.5 / 76.4 / 83.751.4 / 57.6 / 65.5Aug. 20, 2019, 6:34 a.m.
Match Consistency with Fine-Grained Segmentations (GSMC-200)86.4 / 91.2 / 93.877.0 / 82.9 / 88.738.9 / 43.4 / 50.0Aug. 20, 2019, 7:44 p.m.
Localizing Visual Landmarks for 2D matching84.3 / 89.3 / 93.068.0 / 75.1 / 84.442.4 / 51.4 / 69.7April 25, 2019, 8:12 a.m.
Active Search v1.181.0 / 87.3 / 92.462.6 / 70.9 / 81.045.5 / 51.6 / 62.0April 26, 2020, 9:27 a.m.
CityScaleLocalization71.2 / 74.6 / 78.757.8 / 61.7 / 67.534.5 / 37.0 / 42.2March 8, 2019, 8:20 p.m.
Active Search73.1 / 78.0 / 80.352.6 / 59.0 / 63.933.6 / 37.8 / 41.3Dec. 19, 2019, 12:35 p.m.
LocFeatsBaseline53.1 / 62.6 / 88.640.9 / 47.6 / 76.820.4 / 26.3 / 51.3April 15, 2020, 7:12 a.m.
sgVLAD + ELF58.3 / 65.8 / 69.046.5 / 54.2 / 57.916.1 / 19.2 / 20.7May 7, 2020, 5:13 a.m.
LocFeatsPoseLossNotDone50.8 / 57.5 / 89.138.8 / 42.3 / 77.620.6 / 23.5 / 52.2April 15, 2020, 7:13 a.m.
Localizing Visual Landmarks for Place Recognition17.3 / 42.5 / 89.05.8 / 19.4 / 76.16.6 / 23.1 / 73.0June 2, 2019, 10:07 p.m.
DGCNCCC17.1 / 41.5 / 89.18.9 / 26.8 / 77.14.8 / 16.2 / 63.3June 2, 2019, 1:27 p.m.
DASGIL15.2 / 36.1 / 80.56.1 / 19.9 / 74.96.6 / 22.7 / 73.9Feb. 26, 2020, 1:59 a.m.
Netvlad_localization16.4 / 38.5 / 78.98.8 / 27.1 / 77.04.7 / 16.0 / 63.2June 25, 2019, 9:44 p.m.
DIFL-FCL14.8 / 35.1 / 79.65.6 / 18.2 / 69.86.1 / 20.7 / 69.1Feb. 25, 2020, 3:24 a.m.
DenseVLAD14.7 / 36.3 / 83.95.3 / 18.7 / 73.95.2 / 19.1 / 62.0March 12, 2019, 3:21 p.m.
vanilla for DASGIL: no GAN loss12.7 / 31.1 / 74.14.5 / 14.4 / 55.64.4 / 15.2 / 56.0Feb. 29, 2020, 6 a.m.
NetVLAD12.2 / 31.5 / 89.83.7 / 13.9 / 74.72.6 / 10.4 / 55.9June 2, 2019, 9:12 a.m.
wasabi7.9 / 21.3 / 75.23.8 / 13.9 / 67.32.4 / 9.1 / 54.5May 4, 2020, 8:20 a.m.
sgVLAD7.3 / 19.9 / 76.93.6 / 12.8 / 69.12.1 / 9.1 / 57.4May 4, 2020, 8:32 a.m.
joint image segemation and depth to posenet0.3 / 1.5 / 31.90.1 / 0.3 / 12.50.0 / 0.2 / 9.0May 31, 2019, 11:46 a.m.

RobotCar Seasons dataset

Localization thresholds:

  • All conditions: (0.25m, 2°) / (0.5m, 5°) / (5m, 10°)
Methodday allnight allLast updated
KAPTURE-R2D2-APGeM55.1 / 82.1 / 97.328.8 / 58.8 / 89.4May 24, 2020, 3:10 p.m.
Visual Localization Using Dense Semantic 3D Map And Hybrid Features54.6 / 81.9 / 96.914.8 / 33.0 / 51.3May 26, 2020, 4:51 a.m.
RT_AP_IR+CRBNet55.3 / 81.8 / 98.511.5 / 26.5 / 39.2March 6, 2020, 2:38 a.m.
SIFT+IR50+SIG+FM+R70_85+MUL+RE3+GPNP+MERGE57.2 / 81.5 / 97.49.3 / 30.1 / 53.3Sept. 2, 2019, 2:03 a.m.
Geometric Prior Guided Camera Localization57.3 / 81.7 / 97.68.0 / 21.6 / 40.7March 1, 2020, 10:14 a.m.
Visual Localization Using Sparse Semantic 3D Map54.5 / 81.6 / 96.712.3 / 28.5 / 46.5June 2, 2019, 4 a.m.
DenseVLAD & D2-Net54.5 / 80.0 / 95.320.4 / 40.1 / 55.0Oct. 31, 2019, 12:58 a.m.
Active Search v1.1 w/ pose prior, image triplets52.4 / 81.4 / 97.512.6 / 23.2 / 42.0April 10, 2020, 6:28 p.m.
ALL_NV_SP53.9 / 80.4 / 97.58.1 / 21.0 / 35.3April 9, 2020, 2:17 a.m.
S2DHM feature optimization52.8 / 80.0 / 95.234.4 / 69.7 / 94.6May 24, 2020, 12:38 p.m.
Hierarchical-Localization (multi-camera when available)53.8 / 80.4 / 96.011.2 / 27.7 / 49.1June 2, 2019, 6:54 a.m.
ToDayGan, NetVLAD, D2-Net52.2 / 80.1 / 95.920.3 / 46.8 / 73.7Dec. 12, 2019, 7:57 p.m.
Semantic Match Consistency50.3 / 79.3 / 95.27.1 / 22.4 / 45.3March 18, 2019, 11:47 p.m.
Active Search v1.1 w/ pose prior50.9 / 80.2 / 96.66.9 / 15.6 / 31.7April 26, 2020, 9:28 a.m.
superpoint52.3 / 78.9 / 94.315.9 / 37.8 / 87.8Oct. 11, 2019, 9:28 a.m.
Active Search on sequences of camera triplets (uses ground truth relative poses)46.6 / 80.1 / 97.05.8 / 21.0 / 43.1March 13, 2019, 12:03 p.m.
Hierarchical-Localization NetVLAD+SuperPoint53.1 / 79.1 / 95.57.2 / 17.4 / 34.4June 1, 2019, 9:50 a.m.
msnet53.2 / 78.4 / 96.44.0 / 11.3 / 21.3Oct. 31, 2019, 12:38 p.m.
Active Search v1.1, image triplets49.7 / 79.7 / 95.54.7 / 11.4 / 21.9April 10, 2020, 6:30 p.m.
Asymmetric Hypercolumn Matching45.7 / 78.0 / 95.122.3 / 61.8 / 94.5June 5, 2019, 7:25 p.m.
s2dhm46.4 / 77.6 / 95.130.0 / 68.8 / 94.6Oct. 11, 2019, 3:31 a.m.
Localizing Visual Landmarks for 2D matching51.1 / 77.7 / 92.313.8 / 30.3 / 57.7May 16, 2019, 2:09 a.m.
EffecientPairWise2D-3DMatching42.6 / 78.2 / 95.78.5 / 14.1 / 20.6March 20, 2019, 9:52 a.m.
Cascaded Parallel Filtering for Memory-Efficient Image-based Localization48.0 / 78.0 / 94.23.4 / 9.5 / 17.0July 19, 2019, 10:08 a.m.
Active Seach on Camera Triplets45.5 / 77.0 / 94.72.7 / 6.9 / 12.1March 13, 2019, 9:31 a.m.
Active Search v1.143.6 / 76.0 / 94.01.8 / 7.4 / 14.2April 26, 2020, 9:28 a.m.
CityScaleLocalization45.3 / 73.5 / 90.10.6 / 2.6 / 7.2March 13, 2019, 11:59 a.m.
Active Search35.6 / 67.9 / 90.40.9 / 2.1 / 4.3March 13, 2019, 9:04 a.m.
NetVLAD+SSM(stage II)10.1 / 39.7 / 95.95.4 / 21.6 / 71.0April 24, 2020, 10:06 p.m.
SSM10.1 / 39.5 / 94.25.1 / 20.2 / 62.0April 25, 2020, 7:36 a.m.
DenseVLAD (single-scale, top-1 interpolation)9.8 / 32.8 / 90.71.8 / 6.6 / 25.6May 29, 2019, 3:27 p.m.
DenseVLAD (single-scale)8.1 / 31.8 / 90.71.3 / 5.7 / 25.5May 29, 2019, 1 p.m.
VPRI2I8.5 / 31.3 / 85.84.1 / 14.2 / 61.0March 25, 2020, 10:25 a.m.
ToDayGAN + DenseVLAD7.6 / 31.2 / 91.22.2 / 10.8 / 50.5May 27, 2019, 8:05 p.m.
DenseVLAD7.6 / 31.2 / 91.21.0 / 4.4 / 22.7April 11, 2019, 5:44 p.m.
Localizing Visual Landmarks for Place Recognition7.9 / 30.0 / 85.94.1 / 15.7 / 59.1June 2, 2019, 10:07 p.m.
DenseVLAD (single scale) with CLAHE pre-processing7.4 / 30.1 / 88.70.9 / 5.0 / 25.3Aug. 23, 2019, 12:37 p.m.
coarse2fine8.4 / 30.0 / 72.23.4 / 8.7 / 25.1April 2, 2020, 12:17 p.m.
NetVLAD6.4 / 26.3 / 90.90.3 / 2.3 / 15.9June 2, 2019, 9:12 a.m.
DIFL+FCL7.6 / 26.2 / 75.92.5 / 6.5 / 15.8April 3, 2020, 3:34 a.m.
DGCNCCC7.2 / 28.3 / 87.60.0 / 1.1 / 13.4May 21, 2019, 7:38 a.m.
nearest_neighbor4.1 / 17.9 / 87.02.4 / 11.6 / 84.6Oct. 11, 2019, 9:09 a.m.
COCL_HR12.7 / 60.6 / 75.40.5 / 1.6 / 2.2May 30, 2019, 9:20 a.m.
FAB-MAP2.7 / 11.8 / 37.30.0 / 0.0 / 0.0March 13, 2019, 11:50 a.m.
joint image segemation and depth to posenet0.1 / 0.8 / 31.00.0 / 0.0 / 0.0May 31, 2019, 11:46 a.m.
baseline_8000.0 / 0.0 / 0.05.2 / 16.5 / 34.4Nov. 9, 2019, 7:12 a.m.

InLoc dataset

Localization thresholds:

  • All conditions: (0.25m, 10°) / (0.5m, 10°) / (1m, 10°)
Methodduc1duc2Last updated
perl-nvsg50.0 / 69.7 / 78.354.2 / 72.5 / 80.2May 21, 2020, 12:34 a.m.
Hierarchical-Localization + SuperGlue49.0 / 69.2 / 79.853.4 / 77.1 / 80.9May 25, 2020, 6:05 a.m.
ONavi41.9 / 68.2 / 84.350.4 / 76.3 / 80.2May 24, 2020, 12:32 p.m.
DensePE+Sparse-NCNet+DensePV47.0 / 67.2 / 79.843.5 / 64.9 / 80.2May 13, 2020, 3:03 p.m.
Visual Localization Using Dense Semantic 3D Map And Hybrid Features48.0 / 62.6 / 79.353.4 / 64.1 / 74.8May 25, 2020, 6:26 a.m.
Oracle (InLoc++)43.9 / 66.2 / 78.343.5 / 63.4 / 76.3March 4, 2020, 7:47 a.m.
DensePE+NCNet+DensePV43.4 / 64.6 / 77.845.0 / 62.6 / 73.3May 13, 2020, 3:03 p.m.
D2-Net - multi-scale (210 poses) & InLoc (146 poses)42.9 / 63.1 / 75.340.5 / 61.8 / 77.9Nov. 1, 2019, 9:50 a.m.
DensePV+S (w/ scan-graph)40.9 / 63.6 / 71.742.7 / 61.8 / 71.0Aug. 20, 2019, 4:39 p.m.
DensePNV+S (w/ scan-graph)40.4 / 62.1 / 73.742.7 / 61.1 / 70.2Aug. 20, 2019, 4:36 p.m.
DensePNV (w/ scan-graph)39.4 / 62.1 / 74.742.7 / 61.1 / 70.2Aug. 20, 2019, 4:51 p.m.
D2-Net - multi-scale off-the-shelf (174 poses) & InLoc (182 poses)43.9 / 61.6 / 73.742.0 / 60.3 / 74.8Nov. 1, 2019, 9:50 a.m.
DensePNV+S (w/o scan-graph)39.9 / 60.1 / 71.240.5 / 60.3 / 74.8Aug. 20, 2019, 4:57 p.m.
InLoc (w/ scan-graph)38.9 / 59.6 / 67.742.0 / 61.1 / 71.0Aug. 20, 2019, 4:45 p.m.
DensePV+S (w/o scan-graph)40.9 / 60.6 / 71.738.2 / 55.7 / 70.2Aug. 20, 2019, 4:47 p.m.
DensePNV (w/o scan-graph)39.9 / 58.6 / 70.740.5 / 60.3 / 74.8Aug. 20, 2019, 4:50 p.m.
Visual Localization Using Sparse Semantic 3D Map41.4 / 59.1 / 71.238.2 / 49.6 / 58.0June 2, 2019, 4 a.m.
InLoc40.9 / 58.1 / 70.235.9 / 54.2 / 69.5April 30, 2019, 1:52 p.m.
D2-Net - multi-scale44.4 / 58.6 / 71.231.3 / 49.6 / 67.9Nov. 1, 2019, 9:50 a.m.
DFC-Net42.9 / 57.1 / 69.232.1 / 51.1 / 67.2Sept. 19, 2019, 12:57 p.m.
D2-Net - multi-scale off-the-shelf42.4 / 56.6 / 66.231.3 / 50.4 / 64.1Nov. 1, 2019, 9:50 a.m.
INLOC_PE_PV40.4 / 55.6 / 68.235.9 / 46.6 / 64.9Nov. 26, 2019, 12:02 p.m.
INLOC_PE_LPV37.4 / 51.0 / 63.635.1 / 46.6 / 61.1Nov. 26, 2019, 12:43 p.m.
INLOC_PE34.8 / 46.0 / 59.131.3 / 39.7 / 50.4Nov. 26, 2019, 12:01 p.m.
KAPTURE-R2D2-APGeM21.7 / 37.4 / 54.523.7 / 41.2 / 54.2May 24, 2020, 8:43 p.m.
DisLoc+SparsePE23.7 / 34.3 / 49.516.8 / 24.4 / 32.8May 30, 2019, 9:57 a.m.
IR_Q_ONE_ONE15.7 / 22.2 / 30.37.6 / 16.0 / 25.2Nov. 1, 2019, 9:53 a.m.
IRLP_Q_ONE_ONE16.2 / 21.2 / 28.36.1 / 13.7 / 16.0Nov. 14, 2019, 10:53 a.m.
IRP_Q_ONE_ONE_1_614.1 / 17.7 / 21.79.9 / 14.5 / 19.8Nov. 22, 2019, 5:30 a.m.
IRP_Q_ONE_ONE_113.6 / 17.2 / 20.27.6 / 12.2 / 15.3Nov. 22, 2019, 2:21 a.m.
IRL_Q_ONE_ONE11.1 / 16.7 / 21.27.6 / 10.7 / 15.3Nov. 14, 2019, 2:49 a.m.
DenseVLAD0.0 / 1.5 / 5.10.0 / 0.8 / 2.3April 30, 2019, 3:02 p.m.

SILDa Weather and Time of Day dataset

Localization thresholds:

  • All conditions: (0.25m, 2°) / (0.5m, 5°) / (5m, 10°)
MethodeveningsnownightLast updated
NetVLAD (top-50) & D2-Net - multi-scale29.6 / 67.8 / 94.86.0 / 16.4 / 72.325.6 / 51.6 / 79.9Nov. 1, 2019, 9:50 a.m.
NetVLAD (top-50) & D2-Net - single-scale30.0 / 67.8 / 94.41.4 / 10.8 / 67.525.7 / 51.9 / 79.9Sept. 26, 2019, 10:28 a.m.
KAPTURE-R2D2-APGeM31.9 / 66.6 / 92.50.5 / 5.8 / 89.230.5 / 54.2 / 78.5May 24, 2020, 3:11 p.m.
NetVLAD (top-20) & D2-Net - single-scale28.1 / 66.8 / 93.51.7 / 11.1 / 67.324.4 / 49.0 / 75.7Sept. 25, 2019, 3:05 p.m.
DenseVLAD (top-50) + Colmap Image Registrator30.6 / 63.3 / 82.41.9 / 12.0 / 58.221.6 / 44.4 / 69.3Sept. 24, 2019, 8:58 a.m.
Visual Localization Using Sparse Semantic 3D Map29.4 / 66.5 / 93.42.2 / 11.1 / 57.420.4 / 41.8 / 69.0June 2, 2019, 4 a.m.
DenseVLAD (top-20) + Colmap Image Registrator28.2 / 56.2 / 71.23.1 / 9.4 / 36.816.3 / 33.4 / 50.7April 30, 2019, 12:24 p.m.
Localizing Visual Landmarks for Place Recognition0.3 / 7.1 / 41.00.0 / 0.0 / 0.90.0 / 2.7 / 56.9June 2, 2019, 10:06 p.m.
DenseVLAD0.2 / 6.2 / 42.10.0 / 0.0 / 0.70.0 / 2.6 / 53.9April 18, 2019, 12:59 p.m.
DGCNCCC0.0 / 0.0 / 0.00.0 / 0.0 / 0.00.0 / 0.0 / 0.0May 19, 2019, 1:16 p.m.

Symphony Seasons dataset

Localization thresholds:

  • All conditions: (0.25m, 2°) / (0.5m, 5°) / (5m, 10°)
Methodwinter overcastspring sunspring overcastautumn overcastLast updated