Research map / Point clouds / Segmentation and detection
PIXOR: Real-Time 3D Object Detection From Point Clouds
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use PIXOR instead of MV3D | KITTI BEV detection · KITTI val, 0-70m | AP_0.7 (%): 66.32 -> 75.74 | Helped |
| Use PIXOR instead of MV3D | KITTI BEV detection · KITTI test, Moderate | AP_KITTI (%): 77.00 -> 77.05 | Helped |
| Use PIXOR instead of MV3D | KITTI BEV detection · KITTI test, Hard | AP_KITTI (%): 68.94 -> 72.95 | Helped |
| Use PIXOR instead of the customized YOLO-like baseline | TOR4D vehicle detection · TOR4D test | AP_0.7 (%): 69.4% -> 73.3% | Helped |
| Add focal classification loss instead of cross-entropy (ablation) | PIXOR · KITTI val-dev | AP_0.7 (%): 73.46% -> 74.93% | Helped |
| Train directly with decoding loss instead of smooth_L1 regression (ablation) | PIXOR · KITTI val-dev | AP_0.7 (%): 74.93% -> 71.05% | Hurt |
| Fine-tune a smooth_L1-trained model with decoding loss (ablation) | PIXOR · KITTI val-dev | AP_0.7 (%): 74.93% -> 77.16% | Helped |
| Ignore pixels near object boundaries during training (ablation) | PIXOR · KITTI val-dev | AP_0.7 (%): 71.10% -> 74.54% | Helped |
| Add rotation and flip augmentation (ablation) | PIXOR trained with ignored boundary pixels · KITTI val-dev | AP_0.7 (%): 74.54% -> 74.93% | Helped |
| Replace PIXOR backbone with PVANet (ablation) | PIXOR · KITTI val-dev | AP_avg (%): 55.07% -> 51.28% | Hurt |
| Replace PIXOR backbone with ResNet-50 (ablation) | PIXOR · KITTI val-dev | AP_avg (%): 55.07% -> 53.03% | Hurt |
| Replace PIXOR backbone with VGG16-half (ablation) | PIXOR · KITTI val-dev | AP_avg (%): 55.07% -> 54.46% | Hurt |
Closest papers by meaning
- PointPillars: Fast Encoders for Object Detection From Point Clouds (CVPR 2019 · similarity 0.89)
- RangeIoUDet: Range Image Based Real-Time 3D Object Detector Optimized by Intersection Over Union (CVPR 2021 · similarity 0.88)
- Pseudo-LiDAR From Visual Depth Estimation: Bridging the Gap in 3D Object Detection for Autonomous Driving (CVPR 2019 · similarity 0.88)
- Towards Unsupervised Object Detection From LiDAR Point Clouds (CVPR 2023 · similarity 0.87)
- OpenAD: Open-World Autonomous Driving Benchmark for 3D Object Detection (NeurIPS 2025 · similarity 0.86)
- CorrBEV: Multi-View 3D Object Detection by Correlation Learning with Multi-modal Prototypes (CVPR 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.