Research map / Point clouds / Point cloud classification
Lidar Waveforms are Worth 40x128x33 Words
Dominik Scheuble, Hanno Holzhüter, Steven Peters, Mario Bijelic, Felix Heide
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| process neighboring waveforms jointly with a transformer instead of per-waveform peak finding | Lidar DSP · Real clear test set | Chamfer distance (m): 0.626 -> 0.307 | Improved |
| add fine-tuning on real data after simulation (ablation) | Neural DSP · Real clear test set | Chamfer distance (m): 0.535 -> 0.307 | Improved |
| add synthetic pre-training (ablation) | Neural DSP · Real clear test set | Chamfer distance (m): 0.443 -> 0.307 | Improved |
| add high-flux (retroreflector, blooming) simulation to training (ablation) | Neural DSP · Real clear test set | Chamfer distance (m): 0.384 -> 0.307 | Improved |
| add synthetic fog simulation to training (ablation) | Neural DSP · Real fog test set | Chamfer distance (m): 2.311 -> 1.110 | Improved |
| use waveform-based neural DSP vs transient imaging baselines for maximum range | Neural DSP · Real fog data | Maximum range (m): 39.55 -> 56.71 | Improved |
| denoising-style transient imaging methods at high SNR | Lindell and Peng baselines · Real clear test set (SNR 2-4) | Chamfer distance (m): 0.402 -> 0.702 (Lindell), 0.501 (Peng) | Hurt |
| use waveform cues for super-resolution instead of range-image SR | Neural DSP SR · Real SR (a=3) | Chamfer distance (m): 0.413 -> 0.273 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- Tulip — TULIP: Transformer for Upsampling of LiDAR Point Clouds (CVPR 2024) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- TULIP: Transformer for Upsampling of LiDAR Point Clouds (CVPR 2024 · similarity 0.85)
- LidaRF: Delving into Lidar for Neural Radiance Field on Street Scenes (CVPR 2024 · similarity 0.85)
- NeRF-LiDAR: Generating Realistic LiDAR Point Clouds with Neural Radiance Fields (AAAI 2024 · similarity 0.85)
- Polarization Wavefront Lidar: Learning Large Scene Reconstruction from Polarized Wavefronts (CVPR 2024 · similarity 0.85)
- GS-LiDAR: Generating Realistic LiDAR Point Clouds with Panoramic Gaussian Splatting (ICLR 2025 · similarity 0.84)
- Neural Reconstruction of LiDAR Point Clouds under Jamming Attacks via Full-Waveform Representation and Simultaneous Laser Sensing (NeurIPS 2026 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
