mlsanityData to 7 Oct 2026

Structured Knowledge Distillation Towards Efficient and Compact Multi-View 3D Detection

arXiv 2211.08398 ↗First posted 14 Nov 2022cs.CV cs.AIcited by 3

Not on the map. Why?

Abstract (arXiv)

Detecting 3D objects from multi-view images is a fundamental problem in 3D computer vision. Recently, significant breakthrough has been made in multi-view 3D detection tasks. However, the unprecedented detection performance of these vision BEV (bird's-eye-view) detection models is accompanied with enormous parameters and computation, which make them unaffordable on edge devices. To address this problem, in this paper, we propose a structured knowledge distillation framework, aiming to improve the efficiency of modern vision-only BEV detection models. The proposed framework mainly includes: (a) spatial-temporal distillation which distills teacher knowledge of information fusion from different timestamps and views, (b) BEV response distillation which distills teacher response to different pillars, and (c) weight-inheriting which solves the problem of inconsistent inputs between students and teacher in modern transformer architectures. Experimental results show that our method leads to an average improvement of 2.16 mAP and 2.27 NDS on the nuScenes benchmark, outperforming multiple baselines by a large margin.

Key facts. Cited by 3 papers in the map (data through 7 Oct 2026). Builds on 61 of its 68 references that are in the map. First posted on arXiv 14 Nov 2022.

Builds on

61 of 68 references.

References · 61

  1. 2017Attention Is All You Needcited by 59,932
  2. 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 11,979
  3. 2015Distilling the Knowledge in a Neural Networkcited by 10,076
  4. 2019Momentum Contrast for Unsupervised Visual Representation Learningcited by 7,482
  5. 2020End-to-End Object Detection with Transformerscited by 7,384
  6. 2019nuScenes: A multimodal dataset for autonomous drivingcited by 4,485
  7. 2019DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lightercited by 4,243
  8. 2022A ConvNet for the 2020scited by 3,545
  9. 2020Deformable DETR: Deformable Transformers for End-to-End Object Detectioncited by 3,410
  10. 2015Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Codingcited by 2,291
  11. 2014FitNets: Hints for Thin Deep Netscited by 1,619
  12. 2020Big Self-Supervised Models are Strong Semi-Supervised Learnerscited by 1,375
38 more
  1. 2022BEVFormer: Learning Bird's-Eye-View Representation from Multi-Camera Images via Spatiotemporal Transformerscited by 1,213
  2. 2020Center-based 3D Object Detection and Trackingcited by 1,143
  3. 2016Paying More Attention to Attention: Improving the Performance of Convolutional Neural Networks via Attention Transfercited by 1,059
  4. 2019Relational Knowledge Distillationcited by 873
  5. 2019Contrastive Representation Distillationcited by 734
  6. 2021BEVDet: High-performance Multi-camera 3D Object Detection in Bird-Eye-Viewcited by 711
  7. 2015Distillation as a Defense to Adversarial Perturbations against Deep Neural Networkscited by 675
  8. 2021DETR3D: 3D Object Detection from Multi-view Images via 3D-to-2D Queriescited by 655
  9. 2017Deep Mutual Learningcited by 621
  10. 2022BEVDepth: Acquisition of Reliable Depth for Multi-view 3D Object Detectioncited by 568
  11. 2022PETR: Position Embedding Transformation for Multi-View 3D Object Detectioncited by 550
  12. 2019Similarity-Preserving Knowledge Distillationcited by 507
  13. 2019Be Your Own Teacher: Improve the Performance of Convolutional Neural Networks via Self Distillationcited by 491
  14. 2018PACT: Parameterized Clipping Activation for Quantized Neural Networkscited by 456
  15. 2021FCOS3D: Fully Convolutional One-Stage Monocular 3D Object Detectioncited by 444
  16. 2022PETRv2: A Unified Framework for 3D Perception from Multi-Camera Imagescited by 351
  17. 2019A Comprehensive Overhaul of Feature Distillationcited by 341
  18. 2022BEVDet4D: Exploit Temporal Cues in Multi-camera 3D Object Detectioncited by 340
  19. 2019Variational Information Distillation for Knowledge Transfercited by 321
  20. 2017Relation Networks for Object Detectioncited by 305
  21. 2019Correlation Congruence for Knowledge Distillationcited by 280
  22. 20163D Bounding Box Estimation Using Deep Learning and Geometrycited by 277
  23. 2019Structured Knowledge Distillation for Dense Predictioncited by 277
  24. 2021Knowledge distillation: A good teacher is patient and consistentcited by 255
  25. 2019M3D-RPN: Monocular 3D Region Proposal Network for Object Detectioncited by 251
  26. 2017Incremental Network Quantization: Towards Lossless CNNs with Low-Precision Weightscited by 211
  27. 2019Distilling Object Detectors with Fine-grained Feature Imitationcited by 188
  28. 2022Time Will Tell: New Outlooks and A Baseline for Temporal Multi-View 3D Object Detectioncited by 177
  29. 2022Image-to-Lidar Self-Supervised Distillation for Autonomous Driving Datacited by 126
  30. 2022Point-to-Voxel Knowledge Distillation for LiDAR Semantic Segmentationcited by 118
  31. 2020BERT-of-Theseus: Compressing BERT by Progressive Module Replacingcited by 112
  32. 2019Knowledge Adaptation for Efficient Semantic Segmentationcited by 105
  33. 2021Distilling Object Detectors via Decoupled Featurescited by 105
  34. 2018Boosting Self-Supervised Learning via Knowledge Transfercited by 92
  35. 2022BEVStereo: Enhancing Depth Estimation in Multi-view 3D Object Detection with Dynamic Temporal Stereocited by 89
  36. 2021LIGA-Stereo: Learning LiDAR Geometry Aware Representations for Stereo-based 3D Detectorcited by 69
  37. 2022MonoDistill: Learning Spatial Features for Monocular 3D Object Detectioncited by 68
  38. 2018Label Refinery: Improving ImageNet Classification through Label Progressioncited by 65

50 most cited of 61 All 61 (JSON)

Built on by

Cited by year

Citing papers per year: 2024: 1; 2025: 2.

3 papers.

  1. 2024CRKD: Enhanced Camera-Radar Object Detection with Cross-modality Knowledge Distillationcited by 27
  2. 2025TinyBEV: Cross Modal Knowledge Distillation for Efficient Multi Task Bird's Eye View Perception and Planningcited by 0
  3. 2025IMKD: Intensity-Aware Multi-Level Knowledge Distillation for Camera-Radar Fusioncited by 0