mlsanityData to 14 Oct 2026

GSRender: Deduplicated Occupancy Prediction via Weakly Supervised 3D Gaussian Splatting

arXiv 2412.14579 ↗First posted 19 Dec 2024cs.CV cs.AIcited by 6

Not on the map. Why?

Abstract (arXiv)

Weakly-supervised 3D occupancy perception is crucial for vision-based autonomous driving in outdoor environments. Previous methods based on NeRF often face a challenge in balancing the number of samples used. Too many samples can decrease efficiency, while too few can compromise accuracy, leading to variations in the mean Intersection over Union (mIoU) by 5-10 points. Furthermore, even with surrounding-view image inputs, only a single image is rendered from each viewpoint at any given moment. This limitation leads to duplicated predictions, which significantly impacts the practicality of the approach. However, this issue has largely been overlooked in existing research. To address this, we propose GSRender, which uses 3D Gaussian Splatting for weakly-supervised occupancy estimation, simplifying the sampling process. Additionally, we introduce the Ray Compensation module, which reduces duplicated predictions by compensating for features from adjacent frames. Finally, we redesign the dynamic loss to remove the influence of dynamic objects from adjacent frames. Extensive experiments show that our approach achieves SOTA results in RayIoU (+6.0), while also narrowing the gap with 3D- supervised methods. This work lays a solid foundation for weakly-supervised occupancy perception. The code is available at https://github.com/Jasper-sudo-Sun/GSRender.

Key facts. Cited by 6 papers in the map (data through 14 Oct 2026). Builds on 51 of its 57 references that are in the map. First posted on arXiv 19 Dec 2024.

Builds on

51 of 57 references.

References · 51

  1. 2017Attention Is All You Needcited by 60,361
  2. 2021Swin Transformer: Hierarchical Vision Transformer using Shifted Windowscited by 12,139
  3. 20233D Gaussian Splatting for Real-Time Radiance Field Renderingcited by 6,226
  4. 2014Fully Convolutional Networks for Semantic Segmentationcited by 4,859
  5. 2019nuScenes: A multimodal dataset for autonomous drivingcited by 4,523
  6. 2020Deformable DETR: Deformable Transformers for End-to-End Object Detectioncited by 3,422
  7. 2021Plenoxels: Radiance Fields without Neural Networkscited by 1,613
  8. 2014Depth Map Prediction from a Single Image using a Multi-Scale Deep Networkcited by 1,274
  9. 2022BEVFormer: Learning Bird's-Eye-View Representation from Multi-Camera Images via Spatiotemporal Transformerscited by 1,218
  10. 20234D Gaussian Splatting for Real-Time Dynamic Scene Renderingcited by 1,157
  11. 20242D Gaussian Splatting for Geometrically Accurate Radiance Fieldscited by 1,076
  12. 2020Lift, Splat, Shoot: Encoding Images From Arbitrary Camera Rigs by Implicitly Unprojecting to 3Dcited by 1,044
38 more
  1. 2021BEVDet: High-performance Multi-camera 3D Object Detection in Bird-Eye-Viewcited by 713
  2. 2023pixelSplat: 3D Gaussian Splats from Image Pairs for Scalable Generalizable 3D Reconstructioncited by 684
  3. 2022BEVDepth: Acquisition of Reliable Depth for Multi-view 3D Object Detectioncited by 572
  4. 2020Searching Efficient 3D Architectures with Sparse Point-Voxel Convolutioncited by 423
  5. 2023Tri-Perspective View for Vision-Based 3D Semantic Occupancy Predictioncited by 409
  6. 2023Splatter Image: Ultra-Fast Single-View 3D Reconstructioncited by 364
  7. 2021MonoScene: Monocular 3D Semantic Scene Completioncited by 363
  8. 2022BEVDet4D: Exploit Temporal Cues in Multi-camera 3D Object Detectioncited by 341
  9. 2023Occ3D: A Large-Scale 3D Occupancy Prediction Benchmark for Autonomous Drivingcited by 332
  10. 2023SurroundOcc: Multi-Camera 3D Occupancy Prediction for Autonomous Drivingcited by 316
  11. 2023VoxFormer: Sparse Voxel Transformer for Camera-based 3D Semantic Scene Completioncited by 313
  12. 2024GRM: Large Gaussian Reconstruction Model for Efficient 3D Reconstruction and Generationcited by 301
  13. 2023OccFormer: Dual-path Transformer for Vision-based 3D Semantic Occupancy Predictioncited by 293
  14. 2020LMSCNet: Lightweight Multiscale 3D Semantic Completioncited by 188
  15. 2023FB-OCC: 3D Occupancy Prediction based on Forward-Backward View Transformationcited by 165
  16. 2021(AF)2-S3Net: Attentive Feature Fusion with Adaptive Feature Selection for Sparse Semantic Segmentation Networkcited by 164
  17. 2024GaussianFormer: Scene as Gaussians for Vision-Based 3D Semantic Occupancy Predictioncited by 157
  18. 2023SparseBEV: High-Performance Sparse 3D Object Detection from Multi-Camera Videoscited by 150
  19. 2023SelfOcc: Self-Supervised Vision-Based 3D Occupancy Predictioncited by 133
  20. 2023RenderOcc: Vision-Centric 3D Occupancy Prediction with 2D Rendering Supervisioncited by 128
  21. 20203D Sketch-aware Semantic Scene Completion via Semi-supervised Structure Priorcited by 126
  22. 2022Vision-Centric BEV Perception: A Surveycited by 126
  23. 20243D Geometry-aware Deformable Gaussian Splatting for Dynamic View Synthesiscited by 109
  24. 20213D Object Detection for Autonomous Driving: A Surveycited by 106
  25. 2024GaussianFlow: Splatting Gaussian Dynamics for 4D Content Creationcited by 100
  26. 2024SparseOcc: Rethinking Sparse Latent Representation for Vision-Based Semantic Occupancy Predictioncited by 96
  27. 2020AMVNet: Assertion-based Multi-View Fusion Network for LiDAR Semantic Segmentationcited by 87
  28. 2023Behind the Scenes: Density Fields for Single View Reconstructioncited by 63
  29. 2024Gamba: Marry Gaussian Splatting with Mamba for single view 3D reconstructioncited by 56
  30. 2024OPUS: Occupancy Prediction Using a Sparse Setcited by 56
  31. 2024GaussianOcc: Fully Self-supervised and Efficient 3D Occupancy Estimation with Gaussian Splattingcited by 50
  32. 2023FastPillars: A Deployment-friendly Pillar-based 3D Detectorcited by 39
  33. 2024Occupancy as Set of Pointscited by 38
  34. 2023S4C: Self-Supervised Semantic Scene Completion with Neural Fieldscited by 28
  35. 2021DRINet++: Efficient Voxel-as-point Point Cloud Segmentationcited by 23
  36. 2024GaussianBeV: 3D Gaussian Representation meets Perception Models for BeV Segmentationcited by 20
  37. 2024Panoptic-FlashOcc: An Efficient Baseline to Marry Semantic Occupancy with Panoptic via Instance Centercited by 14
  38. 2023LiON: Learning Point-wise Abstaining Penalty for LiDAR Outlier DetectioN Using Diverse Synthetic Datacited by 3

50 most cited of 51 All 51 (JSON)

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 5; 2026: 1.

6 papers.

  1. 2025OptMerge: Unifying Multimodal LLM Capabilities and Modalities via Model Mergingcited by 13
  2. 2025Beyond Frame-wise Tracking: A Trajectory-based Paradigm for Efficient Point Cloud Trackingcited by 4
  3. 2025MinkOcc: Towards real-time label-efficient semantic occupancy predictioncited by 2
  4. 2025GraphGSOcc: Semantic-Geometric Graph Transformer with Dynamic-Static Decoupling for 3D Gaussian Splatting-based Occupancy Predictioncited by 2
  5. 2025Easy3D-Labels: Supervising Semantic Occupancy Estimation with 3D Pseudo-Labels for Automotive Perceptioncited by 2
  6. 2026RCGDet3D: Rethinking 4D Radar-Camera Fusion-based 3D Object Detection with Enhanced Radar Feature Encodingcited by 0