mlsanityData to 14 Oct 2026

PointSplit: Towards On-device 3D Object Detection with Heterogeneous Low-power Accelerators

arXiv 2504.03654 ↗First posted 4 Mar 2025cs.DC cs.AI cs.CVcited by 2

Not on the map. Why?

Abstract (arXiv)

Running deep learning models on resource-constrained edge devices has drawn significant attention due to its fast response, privacy preservation, and robust operation regardless of Internet connectivity. While these devices already cope with various intelligent tasks, the latest edge devices that are equipped with multiple types of low-power accelerators (i.e., both mobile GPU and NPU) can bring another opportunity; a task that used to be too heavy for an edge device in the single-accelerator world might become viable in the upcoming heterogeneous-accelerator world.To realize the potential in the context of 3D object detection, we identify several technical challenges and propose PointSplit, a novel 3D object detection framework for multi-accelerator edge devices that addresses the problems. Specifically, our PointSplit design includes (1) 2D semantics-aware biased point sampling, (2) parallelized 3D feature extraction, and (3) role-based group-wise quantization. We implement PointSplit on TensorFlow Lite and evaluate it on a customized hardware platform comprising both mobile GPU and EdgeTPU. Experimental results on representative RGB-D datasets, SUN RGB-D and Scannet V2, demonstrate that PointSplit on a multi-accelerator device is 24.7 times faster with similar accuracy compared to the full-precision, 2D-3D fusion-based 3D detector on a GPU-only device.

Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Builds on 58 of its 75 references that are in the map. First posted on arXiv 4 Mar 2025.

Builds on

58 of 75 references.

References · 58

  1. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,713
  2. 2014Microsoft COCO: Common Objects in Contextcited by 17,608
  3. 2015Distilling the Knowledge in a Neural Networkcited by 10,324
  4. 2019EfficientNet: Rethinking Model Scaling for Convolutional Neural Networkscited by 6,442
  5. 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,248
  6. 2014Fully Convolutional Networks for Semantic Segmentationcited by 4,859
  7. 2018MobileNetV2: Inverted Residuals and Linear Bottleneckscited by 4,795
  8. 2017PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Spacecited by 4,401
  9. 2020Deformable DETR: Deformable Transformers for End-to-End Object Detectioncited by 3,422
  10. 2016DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFscited by 3,241
  11. 2018Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentationcited by 3,223
  12. 2017ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenescited by 3,117
38 more
  1. 2015Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Codingcited by 2,329
  2. 2018The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networkscited by 2,262
  3. 2017Rethinking Atrous Convolution for Semantic Image Segmentationcited by 2,161
  4. 2015Learning both Weights and Connections for Efficient Neural Networkscited by 2,031
  5. 2019Searching for MobileNetV3cited by 1,994
  6. 2017Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inferencecited by 1,593
  7. 2017VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detectioncited by 1,323
  8. 2019EfficientDet: Scalable and Efficient Object Detectioncited by 1,020
  9. 2018MnasNet: Platform-Aware Neural Architecture Search for Mobilecited by 943
  10. 2019PV-RCNN: Point-Voxel Feature Set Abstraction for 3D Object Detectioncited by 891
  11. 2016Multi-View 3D Object Detection Network for Autonomous Drivingcited by 724
  12. 2019Deep Hough Voting for 3D Object Detection in Point Cloudscited by 648
  13. 2017Frustum PointNets for 3D Object Detection from RGB-D Datacited by 640
  14. 2014Semantic Image Segmentation with Deep Convolutional Nets and Fully Connected CRFscited by 554
  15. 2018Quantizing deep convolutional networks for efficient inference: A whitepapercited by 521
  16. 2020MobileBERT: a Compact Task-Agnostic BERT for Resource-Limited Devicescited by 482
  17. 2019PointPainting: Sequential Fusion for 3D Object Detectioncited by 449
  18. 2020Voxel R-CNN: Towards High Performance Voxel-based 3D Object Detectioncited by 410
  19. 2019Q-BERT: Hessian Based Ultra Low Precision Quantization of BERTcited by 359
  20. 2019Point-Voxel CNN for Efficient 3D Deep Learningcited by 343
  21. 2021An End-to-End Transformer Model for 3D Object Detectioncited by 329
  22. 2021Voxel Transformer for 3D Object Detectioncited by 247
  23. 2021PV-RCNN++: Point-Voxel Feature Set Abstraction With Local Vector Representation for 3D Object Detectioncited by 232
  24. 20183D-SIS: 3D Semantic Instance Segmentation of RGB-D Scanscited by 217
  25. 2021Group-Free 3D Object Detection via Transformerscited by 215
  26. 2016Ternary Weight Networkscited by 204
  27. 2022SoftGroup for 3D Instance Segmentation on Point Cloudscited by 200
  28. 2022Multimodal Token Fusion for Vision Transformerscited by 136
  29. 2019Morphing and Sampling Network for Dense Point Cloud Completioncited by 132
  30. 2016The Importance of Skip Connections in Biomedical Image Segmentationcited by 126
  31. 2020H3DNet: 3D Object Detection Using Hybrid Geometric Primitivescited by 123
  32. 2020ImVoteNet: Boosting 3D Object Detection in Point Clouds with Image Votescited by 119
  33. 2021FCAF3D: Fully Convolutional Anchor-Free 3D Object Detectioncited by 97
  34. 2022Surface Representation for Point Cloudscited by 84
  35. 2020HVNet: Hybrid Voxel Network for LiDAR Based 3D Object Detectioncited by 82
  36. 2020MLCVNet: Multi-Level Context VoteNet for 3D Object Detectioncited by 80
  37. 2019Augmentation for small object detectioncited by 73
  38. 2021Back-tracing Representative Points for Voting-based 3D Object Detection in Point Cloudscited by 67

50 most cited of 58 All 58 (JSON)

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 1; 2026: 1.

2 papers.

  1. 2025ConcreTizer: Model Inversion Attack via Occupancy Classification and Dispersion Control for 3D Point Cloud Restorationcited by 1
  2. 2026[Emerging Ideas] Artificial Tripartite Intelligence: A Bio-Inspired, Sensor-First Architecture for Physical AIcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025StripDet: Strip Attention-Based Lightweight 3D Object Detection from Point Cloud
  2. 2021Deployment of Deep Neural Networks for Object Detection on Edge AI Devices with Runtime Optimization
  3. 2022BED: A Real-Time Object Detection System for Edge Devices
  4. 2022TorchSparse: Efficient Point Cloud Inference Engine
  5. 2026Hardware Utilization and Inference Performance of Edge Object Detection Under Fault Injection
  6. 2026Achieving Near-Zero-Overhead Multi-Model Hierarchical Classification in Real-Time Detection Pipelines
  7. 2024Lotus: learning-based online thermal and latency variation management for two-stage detectors on edge devices
  8. 2022YOLOv6: A Single-Stage Object Detection Framework for Industrial Applications