- Rating
- 3
- Confidence
- 2
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
VEOcc++ maintains persistent voxel features for online semantic occupancy mapping, then refines the completed map using cached observations. It improves mapping accuracy with the local predictor held fixed and demonstrates embedded deployment, though the specific update design and long-term retention need stronger validation.
Critical strengths
-
The fixed-predictor comparison is convincing: Table 2 shows gains of 3.70 IoU and 3.74 mIoU points over VEOcc fusion.
-
The online/offline setup is clear. Refinement reuses existing observations, and Table 6 separates the contributions of different inputs.
-
The Jetson experiments report latency, cache size, and memory growth, providing useful evidence of practical feasibility.
Critical weaknesses
-
Mean feature accumulation already reaches 57.49 mIoU versus 58.40 for the full model (Table 5). A matched learned gate without the reliability prior would better isolate the proposed update's contribution.
-
The updater uses historical average reliability but excludes observation count (Appendix A.3). I'd like repeated-view and conflicting-observation tests to check whether well-supported memory resists incorrect updates.
-
Mapping depends on external poses, and robot experiments use LiDAR-inertial localization. Pose-noise tests are needed to assess how alignment errors affect accumulated features.
-
Quantitative accuracy tests remain ScanNet-based, while longer robot sequences are evaluated qualitatively. Revisit accuracy and accuracy-over-time curves would better support the long-term memory claim.
-
Cached local evidence adds only 0.34 mIoU over probabilities plus map features (Table 6), yet the floor-scale cache occupies 576.50 MiB. A bounded-cache experiment would clarify this trade-off.
Questions for authors
See Weaknesses.
3 comments
Additional Results on Reliability Guidance, Robustness, and Long-Term Mapping
The tables below present the results discussed in our initial response in greater detail. IoU and mIoU are reported in percent.
1. Reliability prior. (Table 5)
The added control retains the learned gate and candidate adaptation while removing the reliability prior.
| Online update | IoU | mIoU |
|---|---|---|
| Learned gate without reliability prior | 66.92 | 58.06 |
| Full model | 67.12 | 58.40 |
The matched gate reaches 66.92/58.06 IoU/mIoU; adding the reliability prior raises this to 67.12/58.40. The gains of 0.20/0.34 IoU/mIoU points reflect the prior's contribution with the learned gate retained.
2. Repeated and conflicting observations. (Appendix C.4, Table 12)
We repeat the same correct observation N times before one incorrect observation. Evaluation uses a shared cohort of 31,449 voxels across 137 validation scenes, correctly classified before the incorrect update under all settings.
| Correct repetitions N | Prediction flip rate after incorrect update (%) ↓ | Mean gate for incorrect update ↓ | Change in ground-truth probability ↑ |
|---|---|---|---|
| 1 | 29.55 | 0.4795 | −0.2122 |
| 2 | 22.01 | 0.4400 | −0.1756 |
| 4 | 14.92 | 0.3967 | −0.1291 |
| 8 | 11.68 | 0.3652 | −0.0995 |
Increasing correct repetitions from one to eight reduces the flip rate from 29.55% to 11.68% and the mean gate from 0.4795 to 0.3652. The ground-truth class probability change moves closer to zero, from −0.2122 to −0.0995. These results indicate better memory stability under conflicting observations.
3. Pose noise robustness. (Appendix C.3, Table 9)
We simulate cumulative pose noise by injecting translational and yaw drift into the camera poses. Mild, moderate, and severe noise correspond to translational drift of 0.5%, 1.0%, and 2.0% of traveled distance, paired with yaw drift of 0.005°, 0.010°, and 0.020° per meter, respectively. Both correction strategies are also simulated.
Each entry reports IoU/mIoU (%). Bold values in parentheses indicate the gains from offline refinement over online mapping (percentage points).
| Noise | Uncorrected: Online | Uncorrected: Offline | Loop correction: Online | Loop correction: Offline | Post-run rebuild: Online | Post-run rebuild: Offline |
|---|---|---|---|---|---|---|
| Clean | 67.12/58.40 | 67.79/59.37 (+0.67/+0.97) | — | — | — | — |
| Mild | 64.65/55.92 | 65.23/56.70 (+0.58/+0.78) | 65.95/57.26 | 66.57/58.02 (+0.62/+0.76) | 66.64/57.92 | 67.29/58.79 (+0.65/+0.87) |
| Moderate | 59.82/50.95 | 60.20/51.46 (+0.38/+0.51) | 63.41/54.63 | 63.90/55.14 (+0.49/+0.51) | 65.60/56.72 | 66.20/57.50 (+0.60/+0.78) |
| Severe | 50.83/41.61 | 51.03/41.86 (+0.20/+0.25) | 57.45/48.45 | 57.68/48.65 (+0.23/+0.20) | 63.02/53.78 | 63.52/54.40 (+0.50/+0.62) |
Both pose correction strategies mitigate the accuracy degradation caused by accumulated pose noise, while offline refinement consistently provides further improvements across all tested noise levels and pose-handling strategies.
4. Long-term mapping accuracy. (Figs. 7–8)
Fig. 7 evaluates accuracy over the observable regions covered by the corresponding fused frames:
| Fused RGB frames | IoU | mIoU |
|---|---|---|
| 5 | 65.01 | 55.80 |
| 10 | 65.89 | 55.75 |
| 15 | 65.75 | 56.09 |
| 20 | 66.25 | 56.94 |
| 25 | 66.67 | 57.84 |
| 30 | 67.12 | 58.40 |
At the frame level, fusing more RGB frames generally improves IoU and mIoU, reaching 67.12% IoU and 58.40% mIoU with 30 frames.
Fig. 8(a,b) groups accuracy by voxel observation frequency:
| Observation count | IoU | mIoU |
|---|---|---|
| 1 | 62.25 | 53.33 |
| 2 | 66.43 | 57.76 |
| 3 | 68.60 | 59.67 |
| 4 | 69.85 | 60.84 |
| 5 | 70.67 | 61.59 |
| 6 | 71.16 | 62.69 |
| 7 | 71.50 | 63.44 |
| 8 | 71.72 | 63.06 |
| 9 | 71.63 | 63.89 |
| 10 | 71.64 | 63.00 |
At the voxel level, accuracy generally improves with the number of observations. This highlights the benefit of repeated multi-view observations for persistent voxel mapping.
5. Local feature cache overhead.
We have provided a detailed discussion of the local feature cache overhead in our initial response.
Official Comment by Authors
Thank you again for your thoughtful and constructive feedback. We have carefully considered your suggestions and made corresponding revisions in both our response and the revised manuscript. We would greatly appreciate hearing whether these updates have adequately addressed your concerns or if any issues remain. We look forward to further discussion and would be happy to provide additional clarification if needed.
Response to Reliability Guidance, Robustness, and Long-Term Mapping
Thank you for your careful review and constructive suggestions. We have uploaded the revised manuscript, with an overview of the changes provided in our comment “Summary of Changes in the Revised Manuscript.” We address your five concerns below.
1. Reliability prior. We add a capacity-matched control retaining the learned gate and candidate adaptation while removing the reliability prior (Table 5). The full model achieves additional gains of 0.20/0.34 IoU/mIoU points, confirming the contribution of explicit reliability guidance beyond learned gating alone.
2. Repeated and conflicting observations. We conduct a controlled experiment with repeated correct observations followed by conflicting evidence (Appendix C.4, Table 12). The results show that accumulated correct observations lead to lower update weights for subsequent incorrect observations, fewer prediction flips, and smaller degradation in ground-truth class probability. This demonstrates that the learned update mechanism adapts to accumulated evidence and improves memory stability.
3. Pose noise robustness. We evaluate three levels of cumulative pose noise with simulated loop correction and post-run reconstruction (Appendix C.3, Table 9). Both correction strategies substantially recover mapping accuracy, while offline refinement consistently provides additional improvements. In practical robotic systems, localization errors can be continuously mitigated through loop closure and factor-graph optimization. Our results suggest that, with such pose maintenance, accumulated drift need not cause the substantial degradation observed under uncorrected noisy poses.
4. Long-term mapping accuracy. We acknowledge that precise quantitative evaluation of real-world mapping accuracy is difficult without dense ground-truth annotations. To provide quantitative evidence, we add accuracy-over-time analysis with accumulated RGB observations and voxel-level accuracy analysis across observation frequencies (Figs. 7–8). Both exhibit favorable accuracy trends as multi-view observations accumulate, supporting the benefits of persistent feature memory during repeated observations.
5. Local feature cache overhead. We appreciate this suggestion. For the 235 m Laboratory Floor sequence, the complete local feature cache occupies only 576.50 MiB, a modest requirement compared with storage capacities commonly reaching hundreds of GB or several TB. Importantly, the cache is only needed for post-exploration refinement and can be discarded once refinement is completed, avoiding permanent accumulation of cached observations. The complete pipeline also runs successfully on a Jetson Orin 16GB. We agree that bounded-cache experiments could further characterize the storage–accuracy trade-off, although this is a secondary consideration to our central investigation of the benefits of reusing accumulated frame-wise evidence. Given the limited revision time, we prioritized additional experiments addressing memory behavior, robustness, and inference efficiency. If you consider a bounded-cache experiment particularly important to their assessment, please let us know, and we will make every effort to provide it.
We sincerely appreciate your valuable feedback, which has helped strengthen the experimental validation and practical analysis of our framework.