Research map / Detection and segmentation / 3D detection and driving
LabelDistill: Label-guided Cross-modal Knowledge Distillation for Camera-based 3D Object Detection
Sanmin Kim, Youngseok Kim, Sihwan Hwang, Hyeonjun Jeong, Dongsuk Kum
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 LiDAR feature and response distillation (ablation) | BEVDepth ResNet50 · nuScenes val | mAP: 33.6 -> 35.4 | Improved |
| add label distillation on top of LiDAR distillation (ablation) | BEVDepth ResNet50 · nuScenes val | mAP: 35.4 -> 37.0 | Improved |
| add feature partitioning on top of LiDAR + label distillation (ablation) | BEVDepth ResNet50 · nuScenes val | mAP: 37.0 -> 37.9 | Improved |
| train label encoder via inverse function approximation instead of plain AutoEncoder (ablation) | LabelDistill · nuScenes val | mAP: 34.9 -> 36.8 | Improved |
| train label encoder with student features (LabelEnc) instead of inverse function approximation (ablation) | LabelDistill · nuScenes val | mAP: 36.8 -> 34.8 | Hurt |
| use a more accurate label encoder (autoencoder mAP 50.2 to 94.1) (ablation) | LabelDistill · nuScenes val | student mAP: 34.0 -> 36.8 | Improved |
| use equal channel ratio for LiDAR, label and image groups (ablation) | feature partitioning · nuScenes val | mAP: 36.6 -> 37.6 (vs 1:3:2); 37.1 vs 3:1:2 | Improved |
| add label distillation for distant objects (over 30m) (ablation) | LabelDistill · nuScenes val, objects beyond 30m | mASE: 0.342 -> 0.270 | Improved |
| add label distillation for distant objects (over 30m) (ablation) | LabelDistill · nuScenes val, objects beyond 30m | mAOE: 0.534 -> 0.531 | No clear effect |
| apply LabelDistill to BEVDepth ConvNeXt-B | BEVDepth · nuScenes test | mAP: 47.5 -> 52.6 | Improved |
Later papers that compared with this method
3 later accepted papers report LabelDistill as a baseline in their results tables; 3 of them report a better number on most of the results they share with it.
- Adaptive-Smooth LiDAR-Camera Knowledge Distillation with Heterogeneous Fusion for Multi-View 3D Object Detection (AAAI 2026 · better on 2 of 2 shared results)
- RayD3D: Distilling Depth Knowledge Along the Ray for Robust Multi-View 3D Object Detection (AAAI 2026 · better on 1 of 1 shared result)
- MemDistill: Distilling LiDAR Knowledge into Memory for Camera-Only 3D Object Detection (ICCV 2025 · better on 4 of 4 shared results)
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.
- BEVFormer v2 — BEVFormer v2: Adapting Modern Image Backbones to Bird's-Eye-View Recognition via Perspective Supervision (CVPR 2023) · 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
- Cross-Modality Knowledge Distillation Network for Monocular 3D Object Detection (ECCV 2022 · similarity 0.92)
- MixSup: Mixed-grained Supervision for Label-efficient LiDAR-based 3D Object Detection (ICLR 2024 · similarity 0.89)
- SimDistill: Simulated Multi-Modal Distillation for BEV 3D Object Detection (AAAI 2024 · similarity 0.88)
- Fine-grained Image-to-LiDAR Contrastive Distillation with Visual Foundation Models (NeurIPS 2024 · similarity 0.88)
- RadarDistill: Boosting Radar-based Object Detection Performance via Knowledge Distillation from LiDAR Features (CVPR 2024 · similarity 0.88)
- SCKD: Semi-Supervised Cross-Modality Knowledge Distillation for 4D Radar Object Detection (AAAI 2025 · similarity 0.88)
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.
