Neural LiDAR Bundle Adjustment
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Robotics Radio. Generated commentary on the latest robotics and control papers.
Rosa: I'm Rosa.
Dev: And I'm Dev, here with Taro.
Rosa: We're looking at the paper: "Neural LiDAR Bundle Adjustment".
Title and authors: Rosa: So we're looking at this paper called Neural LiDAR Bundle Adjustment. It looks like they're tackling a problem where you want to use neural radiance fields for LiDAR data and then tie that into bundle adjustment to get better maps and poses.
Dev: Yeah, it’s interesting because it focuses specifically on how the density of volume sampling affects things in LiDAR NeRFs. They propose a new algorithm called Neural LiDAR Bundle Adjustment, or NeLD-BA, which they designed for optimizing the map and the sensor poses at the same time using this volume sampling idea.
Taro: I'm curious about how they handle that joint optimization part. If you’re trying to optimize both the map and where you are in space simultaneously, that usually introduces some tricky constraints on how you sample things out there.
Rosa: Well, they tackle that by changing the volume sampling strategy significantly. They found out that if you just have fine samples placed after coarse samples, it doesn't work well for getting gradients during pose optimization.
Dev: That’s a key point because it means the gradient term vanishes in the vanilla approach when you try to optimize those poses, which is a big problem for any optimization loop. They suggest placing the fine samples *before* the coarse ones instead, which they claim fully utilizes those fine samples and cuts out some of that range ambiguity.
Taro: So they are essentially saying that for LiDAR data, you need a more specific way to sample because it's not just about filling space evenly. What else are they doing to make the optimization stable?
Rosa: They’re also bounding the sampling intervals themselves. They clamp the ray sampling intervals into a normalized cube, which is this cube C that goes from negative one to positive one in three dimensions. This helps them improve the overall volume sampling density across point clouds where you have those three dee positions specified <ref:2607.04169#pg1>.
Dev: That cube bound seems like a practical way to ensure consistent density regardless of how sparse the point cloud ends up being. But what about the positional encoding part? They had an issue with gradient amplification when using Fourier features in the positional encoding, so they introduced a surrogate gradient for that.
Taro: A surrogate gradient to keep the direction of the gradient but get rid of that huge amplification factor seems like a clever trick to keep things stable during training. Does it actually help them find better solutions?
Title and authors: Rosa: It does, because they claim this modification preserves the gradient direction while removing that amplification component, which prevents the optimization from getting driven away from what should be the global minimum when using large values for those features.
Dev: And they also have this part where they gradually activate those Fourier features and start activating those sensor poses only after a certain frequency, which is k star frequency. It seems like a way to introduce complexity slowly so the system doesn't get overwhelmed early on.
Taro: That sounds like managing the learning speed. So, to summarize what we’ve heard about Neural LiDAR Bundle Adjustment, they are using this new volume sampling strategy with fine samples placed before coarse ones and bounding those sampling intervals in a cube C to improve density, plus they have a surrogate gradient for the positional encoding features to keep the optimization stable.
Rosa: Exactly. And they combine this with a specific loss function that uses L1 loss on estimated ranges, plus a distribution loss for ray termination, and similar losses for both the coarse and fine estimations. This whole setup is designed to match measured ranges against ground truth ranges.
Dev: It’s important to note that they are building the final three dee map by casting rays from those optimized sensor poses using a hierarchical sampling strategy to figure out the termination distances, which we call zmap <ref:2607.04169#pg1>. That’s how they get their final map structure.
Taro: So if you look at the results, they tested this on datasets like Newer College and FusionPortable, and they claim superior performance in point cloud registration compared to methods like HBA and BALM.
Rosa: They show a substantial reduction in the L1-Chamfer distance of raw maps when using NeLD-BA compared to those other methods. On top of that, NeLD-BA outperforms state-of-the-art mapping algorithms on both the F1 score and L1 distance for the rendered maps.
Dev: They also improved the initial poses significantly when compared to FAST LIO2 for trajectory evaluation, which is a big deal if you’re looking at how well this works in real time or over a sequence of measurements.
Taro: It sounds like they’ve made some tangible improvements in accuracy, but I gotta ask about the limitations. What are the practical hurdles they admit they face when using this approach?
Title and authors: Rosa: They do mention that because of that need to normalize point clouds within that cube C, it presents challenges when dealing with very large scale sequences. Plus, as with many NeRFs, these things still require relatively long training times.
Dev: So the main caveat is scalability for long sequences and the computational cost of training. But they did open up their source code for the community if anyone wants to try it out.
Taro: It’s good to know they’re sharing it because getting these complex methods running on real hardware can be a headache. So, we've seen how Neural LiDAR Bundle Adjustment uses tailored volume sampling and other fixes to stabilize gradients, leading to better registration and mapping results on datasets like Newer College and FusionPortable, though it still has training time concerns for big sequences.
Rosa: That’s the gist of the paper. It shows that tailoring the volume sampling specifically for LiDAR ray models is essential for getting those range measurements right during pose optimization.
Dev: And by introducing things like bounding those intervals and using a surrogate gradient, they manage to keep the optimization process from collapsing, even when using powerful positional encoding features.
Taro: It’s a solid piece of work because it addresses the specific mechanics of LiDAR data that RGB methods don't handle well, moving beyond just fitting colors to actually use those explicit range measurements effectively.
Rosa: Yeah, so for anyone interested in how neural fields can be applied to LiDAR, this paper gives us a concrete way to structure the bundle adjustment process around volume sampling.
Dev: That’s what it boils down to: joint optimization of map and poses using this new NeLD-BA algorithm. It’s a step forward from previous work on LiDAR NeRF-BA by focusing on the underlying volume sampling mechanism.
Taro: It suggests that for robust LiDAR mapping, you can't just apply generic deep learning fixes; you have to understand the geometry of the ray constraints themselves.
Rosa: Right. And it’s a good thing they’ve validated this on those newer datasets, showing tangible gains in both raw map distance and final rendered map quality.
Dev: So, for anyone building something that needs accurate LiDAR mapping or odometry, looking into how volume sampling dictates the optimization path in papers like Neural LiDAR Bundle Adjustment is worth checking out.
The paper's summary: Rosa: So, to recap, this paper is about taking those neural radiance fields for LiDAR data and putting them through a bundle adjustment process that specifically focuses on how we sample the volume of space to get the best map and pose estimates together.
Dev: Right. It's not just applying a standard bundle adjustment to a neural field; it’s tailoring the entire volume sampling strategy around how LiDAR rays work, which is crucial for matching those explicit range measurements against what the ground truth should be.
Taro: So, they found that if you sample too coarsely or in the wrong order, your optimization loop just breaks because those gradient terms disappear when you try to adjust your sensor poses.
Rosa: Exactly. They propose a new way of sampling: placing fine samples right before the coarse ones to fully utilize those high-resolution points and cut out that range ambiguity issue.
Dev: And they also bound those sampling intervals by putting them into a normalized cube, which helps keep the volume sampling density consistent, especially when you're dealing with point clouds that have their positions defined in three dimensions.
Taro: That sounds like a practical fix for getting stable gradients when you're trying to optimize something in real-time or over a long sequence.
Rosa: They also introduced this surrogate gradient for the positional encoding part, which allows them to use those high-frequency features without the gradients getting amplified and pushing the optimization away from the right solution.
Dev: That’s smart because it keeps you near that global minimum even when you’re using very large values for those frequency features.
Taro: And they sharpen the constraint on range estimation by proposing a pseudo dirac delta function as a targeted termination distribution, making sure the ray termination probability is one in the first sample after it terminates.
Rosa: So, to put it all together, this Neural LiDAR Bundle Adjustment algorithm uses these tailored sampling techniques and loss functions—like L1 loss on ranges and distribution losses for terminations—to build a more accurate three dee map by optimizing both the sensor poses and the neural field parameters simultaneously.
Dev: It’s not just about getting a pretty rendered image anymore; they claim superior results in point cloud registration against methods like HBA and BALM, showing a significant reduction in raw map distance errors.
Taro: And for trajectory evaluation, they improved the initial poses compared to FAST LIO2, which means starting your navigation process off much better.
Rosa: The limitation they mention is that because of that cube bound on sampling intervals, it can be tricky when you're working with very large scale sequences, and as with any NeRF model, training those things still takes a fair amount of time.
Dev: So the practical takeaway is this: this method gives you a more accurate map and better starting points for navigation using LiDAR data because it’s deeply integrated with how the volume is sampled during the optimization loop.
Taro: This suggests that for building robust systems that need to do accurate mapping and odometry, you have to pay close attention to the underlying geometry of your sensor data constraints, not just relying on generic deep learning fixes.
Rosa: And it’s a good thing they shared their source code; if you want to test this out on your own hardware, it’s available for anyone interested.
The paper's improvements: Rosa: So, to wrap up what we’ve seen about Neural LiDAR Bundle Adjustment, these authors are really focusing on how they made the optimization process much more stable and accurate by tweaking a few key parts of the setup.
Dev: Yeah, they didn't just throw more parameters at it; they changed the fundamental way those rays are sampled to ensure that when you try to adjust your sensor poses, you get meaningful information back.
Taro: They’re using this new volume sampling strategy where they put fine samples before coarse ones, which is a direct fix for the problem where the vanilla hierarchical sampling strategy just doesn't provide good gradients for pose optimization.
Rosa: That means they’re making sure those high-resolution points are fully used and that it cuts out that ambiguity in range estimation right away.
Dev: And they also introduced this specific modification to the positional encoding, a surrogate gradient, which allows them to use those high frequency Fourier features without having the gradients get amplified and throw your optimization off course.
Taro: That stability is important because if you’re running real-time loops or long sequences, you don't want the optimization just to drift into a bad area when using large values for those features.
Rosa: Plus, they have this targeted termination distribution idea—a pseudo dirac delta function—to make sure the ray termination probability is one in the first sample after it stops.
Dev: That sharpens your constraints on range estimation significantly, which helps you get a much tighter fit to the measured ranges when matching them to ground truth.
Taro: It’s like they’re giving your optimization loop a clearer target so it knows exactly what it needs to focus on during those challenging pose adjustments.
Rosa: So, in short, they've improved the method by making the volume sampling smarter and introducing specific mathematical tricks for the encoding and termination distribution to keep everything working smoothly.
Dev: That means when you apply this to actual robot systems or long mapping tasks, you should see much better pose refinement and more consistent results compared to older bundle adjustment methods.
Taro: The implication here is that for complex LiDAR scenes, you don't just need a good neural field; you need a geometrically aware optimization framework built around the physical constraints of how LiDAR data is actually sampled.
Rosa: And they’ve shown this works on real datasets like Newer College and FusionPortable, proving it holds up when you move from theory to actual point clouds.
Dev: So, for anyone looking at building something that needs accurate LiDAR mapping or odometry, this paper suggests that understanding the geometry of the ray constraints during volume sampling is a critical step beyond just using standard neural field techniques.
Conclusion: Rosa: So we’ve covered how Neural LiDAR Bundle Adjustment uses tailored volume sampling and specific gradient tricks to stabilize the map and pose optimization process, leading to better results on datasets like Newer College and FusionPortable.
Dev: It really boils down to making sure that when you’re trying to optimize your system in real-time or over a long sequence, the algorithm doesn't just get stuck because those gradient terms vanish.
Taro: What I find interesting is how it handles the world misbehaving; this method seems to recover much more gracefully from initial pose errors than some of the standard methods we’ve seen.
Rosa: Exactly. It produces a more accurate pose refinement, which means your trajectory evaluation starts off much better, and for a field robot, that translates to less time spent fixing things on the ground after deployment.
Dev: And from a controls standpoint, I like that it gives you those better initial poses; it shortens the time needed for the control system to settle into a stable loop.
Taro: It’s not just about getting a pretty map; it’s about having a robust odometry pipeline that handles real-world noise and changes well, which is exactly what we need when deploying these systems outside the lab.
Rosa: This paper, "Neural LiDAR Bundle Adjustment," shows how deep geometric awareness in the sampling process can lead to tangible gains in both raw map accuracy and the quality of the final rendered three dee scene.
Dev: It proves that you can use neural fields for LiDAR data without sacrificing the precision you get from traditional bundle adjustment techniques.
Taro: It suggests that for future work, we should look at how they handle very large scale sequences, because they mentioned that bounding those sampling intervals by a normalized cube can be a challenge when dealing with massive point clouds.
Rosa: Yeah, scalability for long-range mapping is definitely something to keep an eye on; it’s not quite perfect for the biggest problems yet.
Dev: Right. So the next thing we should probably look at is how this compares to other methods like RaDiVe or those that use different types of volume sampling strategies, and whether these gains hold up when things get really complex.
Taro: I agree. It’s a solid step forward in making these neural representations more reliable for real-world autonomy, and it opens up new ways to constrain the optimization process itself.
Imperial College London
cs.RO
Submitted: 2026-07-05
Updated: 2026-11-18
Comments: 8 pages, 8 figures, 4 tables
Code: https://github.com/MichaelGrupp/evo
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Importance score: 80/100
The gist: Recent research has achieved remarkable novel view rendering and scene reconstruction results with Neural Radiance Field (NeRF), including extensions to the LiDAR modality.
Key concepts
- Neural LiDAR Bundle Adjustment (NeLD-BA)
- A new algorithm designed to simultaneously optimize the spatial arrangement of a 3D LiDAR map and the precise poses of the sensor. It uses efficient volume sampling of LiDAR rays as a core component, allowing for joint optimization that improves both mapping accuracy and pose refinement.
- Volume Sampling Strategy Modification
- The original hierarchical sampling method was improved by placing fine samples before coarse ones to better utilize high-resolution data and eliminate range ambiguity. Additionally, bounding ray sampling intervals within a normalized cube [-1, 1]³ increases the overall volume sampling density for more accurate optimization gradients.
- Surrogate Gradient for Fourier Positional Encoding
- This technique modifies how high-frequency positional features are used in the neural network. It preserves the direction of gradients while removing amplification effects, preventing pose optimization from being driven away from optimal solutions when using large loss values.
- Range Loss as Primary Cue
- Unlike methods relying only on photometric losses (like RGB NeRF), this system prioritizes the L1 loss on estimated range measurements. This leverages the strong geometric cues from LiDAR rays to refine poses based on explicit distance measurements, leading to more accurate LiDAR pose estimation.
Terminology
Summary
Recent research has achieved remarkable novel view rendering and scene reconstruction results with Neural Radiance Field (NeRF), including extensions to the LiDAR modality. p. 0 This work proposes a novel Neural LiDAR Bundle Adjustment (NeLD-BA) algorithm tailored for joint optimization of LiDAR map and poses based on the density of volume sampling. p. 0
Problem Formulation
The LiDAR NeRF-BA problem is formulated by minimizing a loss function over sensor poses and MLP parameters to match measured ranges to ground truth ranges. The optimization goal is defined as P∗, Θ∗ = arg min P,Θ L(Dˆ D). p. 1
Volume Sampling Strategy
The paper finds that the density of volume sampling is essential for ensuring the gradient term does not vanish during pose optimization. Two modifications are proposed to the volume sampling strategy:
-
Fine samples are placed at volumes before coarse samples, which fully utilizes fine samples and significantly eliminates the estimated range ambiguity. p. 2
-
Each ray’s sampling interval is bounded to a normalized cube C = [-1, 1]3 to improve overall volume sampling density. p. 2
Positional Encoding Enhancement
To handle the gradient amplification issue, a surrogate gradient for Fourier positional encoding is proposed to preserve the gradient direction while removing the amplification component. Additionally, BARF is followed in gradually activating Fourier features and sensor poses starts only when the k∗ th frequency for position features is activated. p. 3
Loss Functions and Mapping
The final loss function L combines L1 loss on estimated range, a distribution loss for ray termination distribution, and similar losses for coarse and fine estimations. A rendered 3D map is built by casting rays from the optimized sensor poses using a hierarchical sampling strategy to determine termination distances zmap. p. 3
Experimental Validation
Extensive experiments using the Newer College and FusionPortable datasets demonstrate superior performance in multi-view point cloud registration and 3D mapping. The method produces a more accurate pose refinement, evident from the substantial reduction in the L1-Chamfer distance of raw maps compared to HBA and BALM. p. 4 Furthermore, NeLD-BA outperforms state-of-the-art mapping algorithms on F1 score and L1-CD for rendered maps. p. 5 The method has substantially improved the initial poses from FAST-LIO2 for trajectory evaluation. p. 5
Limitations
The algorithm faces challenges with large-scale sequences due to the need to normalize point clouds within a [-1, 1]3 cube. Additionally, NeRFs require relatively long training times. p. 6 The authors release the source code for the community upon acceptance of the paper. p. 0
References
[1] B. Mildenhall, P. p. 7 P. Srinivasan, M. p. 7 Tancik, J. T. Barron, R. Ramamoorthi, and R. p. 7 Ng, “Nerf: Representing scenes as neural radiance fields for view synthesis,” Communications of the ACM. p. 7
[2] C.-H. Lin, W.-C. Ma, A. Torralba, and S. p. 7 Lucey, “Barf: Bundleadjusting neural radiance fields,” in Proceedings of the IEEE/CVF International Conference on Computer Vision. p. 7
[3] Z. Wang, S. Wu, W. Xie, M. Chen, and V. p. 7 A. Prisacariu, “Nerf–: Neural radiance fields without known camera parameters,” 2021. p. 7
[4] Y. Chen et al. p. 7 p. 7, “Local-to-global registration for bundle-adjusting neural radiance fields,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. p. 7
[5] X. Liu, Z. Liu, F. Kong, and F. p. 7 Zhang, “Large-scale lidar consistent mapping using hierarchical lidar bundle adjustment,” IEEE Robotics and Automation Letters. p. 7
[6] M. Ramezani, Y. p. 7 Wang, M. Camurri, D. p. 7 Wisth, M. Mattamala, and M. p. 7 Fallon, “The newer college dataset: Handheld lidar, inertial and vision with ground truth,” in 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). p. 7 IEEE.
[7] J. Zhang, S. Singh et al. p. 7, “Loam: Lidar odometry and mapping in real-time,” in Robotics: Science and Systems, vol. p. 7 2, no. 9.
[8] P. J. Besl and N. p. 7 D. McKay, “Method for registration of 3-d shapes,” in Sensor fusion IV: control paradigms and data structures, vol. p. 7 1611.
[9] Y. Chen and G. p. 7 Medioni, “Object modelling by registration of multiple range images,” Image and Vision Computing, vol. p. 7 10, no. 3.
[10] H. Wang et al. p. 0, “Robust multiview point cloud registration with reliable pose graph initialization and history reweighting,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023. p. 7
[11] Z. Qin et al., “Geotransformer: Fast and robust point cloud registration with geometric transformer,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. p. 7 45, no. 8.
[12] F. Lu et al., “Hregnet: A hierarchical network for large-scale outdoor lidar point cloud registration,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2021. p. 7
[13] M. Ruhnke, R. Kümmerle, G. p. 7 Grisetti, and W. p. 7 Burgard, “Highly accurate 3D surface models by sparse surface adjustment,” in IEEE International Conference on Robotics and Automation. p. 7
[14] Z. Liu and F. Zhang, “Balm: Bundle adjustment for lidar mapping,” IEEE Robotics and Automation Letters, vol. p. 7 6, no. 2.
[15] Z. Zheng, F. Lu, W. Xue, G. Chen, and C. p. 7 Jiang, “Lidar4d: Dynamic neural fields for novel space-time view lidar synthesis,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024. p. 7
[16] J. Jiang, C. Gu, Y. Chen, and L. p. 7 Zhang, “Gs-lidar: Generating realistic lidar point clouds with panoramic gaussian splatting,” arXiv preprint arXiv:2501. p. 713971, 2025. p. 7
[17] J. Deng et al. p. 7, “Nerf-loam: Neural implicit representation for largescale incremental lidar odometry and mapping,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2023. p. 7
[18] S. Isaacson, P. p. 7-C. Kung, M. Ramanagopal, R. p. 7 Vasudevan, and K. p. 7 A. Skinner, “Loner: Lidar only neural representations for real-time slam,” IEEE Robotics and Automation Letters, vol. p. 7 8, no. 12.
[19] W. Xue, Z. Zheng, F. Lu, H. Wei, G. Chen et al., “Geonlf: Geometry guided pose-free neural lidar fields,” Advances in Neural Information Processing Systems, vol. p. 7 37.
[20] X. Zhong, Y. Pan, J. Behley, and C. p. 7 p. 7 Stachniss, “Shine-mapping: Large-scale 3d mapping using sparse hierarchical implicit neural representations,” arXiv preprint arXiv:2210. p. 702299, 2022. p. 7
[21] Y. Pan, X. Zhong, L. Wiesmann, T. p. 7 Posewsky, J. p. 7 Behley, and C. Stachniss, “Pin-slam: Lidar slam using a point-based implicit neural representation for achieving global map consistency,” IEEE Transactions on Robotics, vol. p. 7 40.
[22] T. Takikawa et al. p. 7, “Neural geometric level of detail: Real-time rendering with implicit 3d shapes,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2021. p. 7
[23] W. Bian, Z. Wang, K. Li, J.-W. Bian, and V. p. 7 A. Prisacariu, “Nope-nerf: Optimising neural radiance field with no pose prior,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023. p. 7
[24] P. Wang, L. Zhao, R. Ma, and P. Liu, “Bad-nerf: Bundle adjusted deblur neural radiance fields,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023. p. 7
[25] J. T. Barron, B. Mildenhall, M. p. 7 Tancik, P. Hedman, R. Martin-Brualla, and P. p. 7 P. Srinivasan, “Mip-nerf: A multiscale representation for antialiasing neural radiance fields,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2021. p. 7
[26] J. T. Barron, B. Mildenhall, D. p. 7 Verbin, P. P. Srinivasan, and P. p. 7 Hedman, “Zip-nerf: Anti-aliased grid-based neural radiance fields,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2023. p. 7
[27] D. Verbin et al. p. 7, “Nerf-casting: Improved view-dependent appearance with consistent reflections,” in SIGGRAPH Asia 2024 Conference Papers, 2024. p. 7
[28] A. Williams, S. p. 7 Barrus, R. K. Morley, and P. p. 7 Shirley, “An efficient and robust ray-box intersection algorithm,” in ACM SIGGRAPH 2005 Courses, 2005. p. 7
[29] M. Tancik et al. p. 7, “Fourier features let networks learn high frequency functions in low dimensional domains,” Advances in Neural Information Processing Systems, vol. p. 7 33.
[30] R. Martin-Brualla, N. p. 7 Radwan, M. S. Sajjadi, J. p. 7 T. Barron, A. Dosovitskiy, and D. p. 7 Duckworth, “Nerf in the wild: Neural radiance fields for unconstrained photo collections,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021. p. 7
[31] F. Zenke and T. p. 7 P. Vogels, “The remarkable robustness of surrogate gradient learning for instilling complex function in spiking neural networks,” Neural Computation, vol. p. 7 33, no. 4.
[32] W. Xu, Y. Cai, D. He, J. Lin, and F. p. 7 Zhang, “Fast-lio2: Fast direct lidar-inertial odometry,” IEEE Transactions on Robotics, vol. p. 7 38.
[33] K. Deng, A. Liu, J.-Y. Zhu, and D. p. 7 Ramanan, “Depth-supervised nerf: Fewer views and faster training for free,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022. p. 7
[34] A. Paszke et al. p. 7, “Pytorch: An imperative style, high-performance deep learning library,” 2019. p. 7
[35] D. P. Kingma and J. p. 7 Ba, “Adam: A method for stochastic optimization,” arXiv preprint arXiv:1412. p. 76980, 2014. p. 7
[36] X. Zhong, Y. Pan, C. Stachniss, and J. p. 7 Behley, “3d lidar mapping in dynamic environments using a 4d implicit neural representation,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024. p. 7
[37] H. Wei et al., “Fusionportablev2: A unified multi-sensor dataset for generalized slam across diverse platforms and scalable environments,” The International Journal of Robotics Research, vol. p. 7 44, no. 7.
[38] M. Grupp, “evo: Python package for the evaluation of odometry and slam,” https://github. p. 7com/MichaelGrupp/evo, 2017. p. 7
Table I: Evaluation of Raw Point Cloud Maps Using L1-Chamfer Distance [cm]
[38] Newer College FusionPortable Method cloister math easy math hard quad easy quad mid quad hard canteen day canteen night garden day garden night
Table II: Evaluation of Rendered Maps
[37] Newer College FusionPortable L1-CD PIN SLAM SHINE-Mapping 4dNDF NeLD-BA Ours cloister math easy math hard quad easy quad mid quad hard canteen day canteen night garden day garden night
Table III: ATE Mean Values for Trajectory Evaluation for FusionPortable Dataset [meter]
[38] FAST-LIO2 SGHR GeoTrans HBA BALM NeLD-BA Ours canteen day canteen night garden day garden night
Table IV: Ablation Study on L1-Chamfer Distance of Raw Point Cloud Maps [cm]
[38] w/o Hier. Sampling w/o Cube Bound w/o Termination Loss w/o Surrogate Grad Ours cloister math hard quad hard canteen day
Fig. 8: Ablation Study Qualitative Results
[38] NeRF’s [1] sampling method result in an outlier frame (left) while our method successfully registered the outlier frame (right).
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
volume sampling strategy tailored to LiDAR ray model: This improvement ensures that
the density of volume sampling plays a significant role in LiDAR NeRF,
leading to more effective constraint of range measurements during pose optimization. -
introduction of a novel Neural LiDAR Bundle Adjustment (NeLD-BA) algorithm: This algorithm is tailored using
efficient volume sampling of LiDAR rays for joint optimization of LiDAR map and poses,
enabling it to outperform existing methods in both novel view synthesis and odometry. -
incorporation of range loss as the primary optimization cue: Unlike RGB NeRF-BA, this system makes
full use of the range loss due to the strong geometric cue from LiDAR rays,
which is crucial for refining LiDAR poses based on explicit measurements rather than relying solely on photometric losses. -
modified volume sampling strategy for gradient stability: By placing fine samples
at volumes before coarse samples,
the system addresses the issue wherethe vanilla hierarchical sampling strategy [1] is sub-optimal for providing gradients for optimising p,
ensuring a non-vanishing partial derivative in equation 5. -
implementation of a surrogate gradient for Fourier positional encoding: This modification allows for
using a high-frequency Fourier features with no amplification on gradients,
which ensures that the optimization of poses does not get driven away from the global minima, even when using large L values. -
use of a pseudo dirac delta function as the targeted termination distribution: By proposing this as
the targeted termination distribution,
the system effectively provides thatthe ray termination probability in the first sample after ray termination is 1,
which significantly sharpens constraints on range estimation. -
bounding of volume sampling intervals to a normalized cube: The system clamps sampling intervals based on intersections with the cube C, which
significantly improves the overall volume sampling density
and is applicable for point clouds where 3D positions are provided.
Sources
- GS-LiDAR: Generating Realistic LiDAR Point Clouds with Panoramic Gaussian Splatting
- SHINE-Mapping: Large-Scale 3D Mapping Using Sparse Hierarchical Implicit Neural Representations
- PyTorch: An Imperative Style, High-Performance Deep Learning Library
- Adam: A Method for Stochastic Optimization
Related papers
- Rethinking Actions: Continuity-Aware Priors for World-Model Planning
- MOSAIC: Modular Obstacle and Spatial-Aware Framework for Manipulation In Cluttered Scenes
- SAGAS: Semantic-Aware Graph-Assisted Stitching for Offline Temporal Logic Planning
- DexSeed: Expanding Human Hand-Object Interactions into Robot Demonstrations for Dexterous Policy Learning
- LaST0.5: A Generalist World Action Model with Interaction-Centric Latent Reasoning
- CIALLO: Learning Shared Intent Representations for Cross-Embodiment Manipulation