Research map / Label-efficient and robust learning / OOD and anomaly detection
Neural Distribution Prior for LiDAR Out-of-Distribution Detection
Zizhao Li, Zhengkang Xiang, Jiayang Ao, Feng Liu, Joseph West, Kourosh Khoshelham
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 |
|---|---|---|---|
| add Soft Outlier Exposure (SOE) on void points (ablation) | NDP-EE · STU val | AP: 67.10 -> 74.24 | Improved |
| add NDP reweighting on top of static extended energy (ablation) | NDP-EE · STU val | AP: 58.69 -> 74.24 | Improved |
| add Perlin noise OOD samples to void training (ablation) | NDP-EE · STU val | AP: 2.09 -> 11.03 | Improved |
| train on Perlin samples alone and add SOE void regions instead of void-only (ablation) | NDP-EE · STU val | AP: 2.09 -> 67.10 | Improved |
| add learnable NDP to energy score (ablation) | Energy score · STU val | AP: 56.88 -> 66.54 | Improved |
| learnable NDP instead of statistical class prior (ablation) | Energy score · STU val | AP: 23.87 -> 66.54 | Improved |
| learnable NDP instead of Balanced Energy (ablation) | Energy score · STU val | AP: 57.05 -> 66.54 | Improved |
| vary SOE soft label beta from 0.7 to 1.0 (ablation) | NDP-EE · STU val | AP: 68.00 / 72.73 / 74.24 / 73.06 | Mixed |
| NDP embedding size d 16 vs 8 and 32 (ablation) | NDP-EE · STU val | AP: 70.29 / 74.24 / 70.14 | Mixed |
| add NDP with lightweight MinkUNet backbone (ablation) | NDP-EE · STU val (MinkUNet) | AP: 58.36 -> 70.29 | Improved |
| add NDP training to Mask4Former, in-distribution segmentation | Mask4Former · STU val | PQ: 52.73 -> 52.37 | No clear effect |
| NDP-EE instead of NDP-Entropy score | NDP · STU test | AP: 29.53 -> 61.31 | Improved |
Closest papers by meaning
- OLiDM: Object-aware LiDAR Diffusion Models for Autonomous Driving (AAAI 2025 · similarity 0.87)
- Leveraging Perturbation Robustness to Enhance Out-of-Distribution Detection (CVPR 2025 · similarity 0.87)
- LiON: Learning Point-Wise Abstaining Penalty for LiDAR Outlier DetectioN Using Diverse Synthetic Data (AAAI 2025 · similarity 0.87)
- An Empirical Study of the Generalization Ability of Lidar 3D Object Detectors to Unseen Domains (CVPR 2024 · similarity 0.86)
- Detecting Out-of-Distribution Through the Lens of Neural Collapse (CVPR 2025 · similarity 0.86)
- RUNA: Object-Level Out-of-Distribution Detection via Regional Uncertainty Alignment of Multimodal Representations (AAAI 2025 · similarity 0.86)
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.
