| Methods | ATE↓ [cm] |
ATE↓ w/o align [cm] |
Track.↓ [ms x it] |
Map.↓ [ms x it] |
Track. FPS↑ |
Map. FPS↑ |
#param↓ |
|---|---|---|---|---|---|---|---|
| iMAP* | 36.67 | - | 30.4x50 | 44.9x300 | 0.66 | 0.07 | 0.2M |
| NICE-SLAM | 9.63 | 23.97 | 12.3x50 | 125.3x60 | 1.63 | 0.13 | 10.3M |
| Vox-Fusion | 8.22 | - | 29.4x30 | 85.8x15 | 1.13 | 0.78 | 1.1M |
| ESLAM | 7.42 | - | 7.4x30 | 22.4x30 | 4.54 | 1.49 | 10.5M |
| Co-SLAM | 9.37 | 18.01 | 7.8x10 | 20.2x10 | 12.82 | 4.95 | 0.8M |
| Co-SLAM† | 8.75 | - | 7.8x20 | 20.2x10 | 6.41 | 4.95 | 0.8M |
For ScanNet scenes, we found that sometimes ATE RMSE with SE(3) alignment do not fully reflect the quality of tracking result: In the example below (black: GT, red: estimated) the results on the left clearly has larger error drift than the one on the right, yet they have very close ATE RMSE: 8.64 vs 7.18. Thus we also report the ATE RMSE with the raw estimated trajectory without doing SE(3) alignment for a reference.
The runtime is measured as the total processing time from the program starts until the final checkpoint is saved. Mesh saving was disabled for NICE-SLAM and iMAP* as we observed that they can take up to ~10 minutes to extract a single mesh.
| Methods | ATE↓ [cm] |
ATE↓ w/o align [cm] |
runtime↓ | #param↓ |
|---|---|---|---|---|
| iMAP* | 55.95 | - | 6hr33min | 0.20M |
| NICE-SLAM | 8.64 | 25.24 | 2hr46min | 22.04M |
| Vox-Fusion | 8.39 | - | 2hr15min | 1.18M |
| ESLAM | 7.32 | - | 40min10s | 17.8M |
| Co-SLAM | 7.18 | 12.94 | 14min35s | 1.59M |
| Co-SLAM† | 7.13 | - | 20min40s | 1.59M |
| Methods | ATE↓ [cm] |
ATE↓ w/o align [cm] |
runtime↓ | #param↓ |
|---|---|---|---|---|
| iMAP* | 32.06 | - | 2hr10min | 0.20M |
| NICE-SLAM | 12.25 | 25.01 | 46min34s | 6.88M |
| Vox-Fusion | 9.18 | - | 44min20s | 1.13M |
| ESLAM | 8.55 | - | 12min24s | 8.40M |
| Co-SLAM | 12.29 | 19.12 | 4min25s | 0.13M |
| Co-SLAM† | 11.14 | - | 6min30s | 0.13M |
| Methods | ATE↓ [cm] |
ATE↓ w/o align [cm] |
runtime↓ | #param↓ |
|---|---|---|---|---|
| iMAP* | 17.50 | - | 2hr50min | 0.20M |
| NICE-SLAM | 8.09 | 10.40 | 1hr2min | 11.07M |
| Vox-Fusion | 7.44 | - | 1hr2min | 1.10M |
| ESLAM | 7.51 | - | 15min31s | 11.2M |
| Co-SLAM | 9.57 | 12.12 | 5min50s | 0.48M |
| Co-SLAM† | 9.36 | - | 8min55s | 0.48M |
| Methods | ATE↓ [cm] |
ATE↓ w/o align [cm] |
runtime↓ | #param↓ |
|---|---|---|---|---|
| iMAP* | 70.51 | - | 2hr25min | 0.20M |
| NICE-SLAM | 10.28 | 30.51 | 49min55s | 7.14M |
| Vox-Fusion | 6.53 | - | 49min50s | 1.03M |
| ESLAM | 6.51 | - | 13min37s | 8.58M |
| Co-SLAM | 6.62 | 19.61 | 4min52s | 1.59M |
| Co-SLAM† | 5.90 | - | 7min40s | 1.59M |
| Methods | ATE↓ [cm] |
ATE↓ w/o align [cm] |
runtime↓ | #param↓ |
|---|---|---|---|---|
| iMAP* | 32.10 | - | 2hr48min | 0.20M |
| NICE-SLAM | 12.93 | 39.98 | 1hr | 6.90M |
| Vox-Fusion | 12.20 | - | 46min | 1.05M |
| ESLAM | 9.21 | - | 15min57s | 8.48M |
| Co-SLAM | 13.43 | 34.41 | 5min35s | 0.48M |
| Co-SLAM† | 11.81 | - | 8min20s | 0.48M |
| Methods | ATE↓ [cm] |
ATE↓ w/o align [cm] |
runtime↓ | #param↓ |
|---|---|---|---|---|
| iMAP* | 11.91 | - | 2hr23min | 0.20M |
| NICE-SLAM | 5.59 | 12.70 | 49min5s | 7.50M |
| Vox-Fusion | 5.57 | - | 44min | 1.10M |
| ESLAM | 5.71 | - | 13min34s | 8.50M |
| Co-SLAM | 7.13 | 9.88 | 4min40s | 0.48M |
| Co-SLAM† | 7.14 | - | 7min24s | 0.48M |