mlsanityData to 7 Oct 2026

SelfOccFlow: Towards end-to-end self-supervised 3D Occupancy Flow prediction

arXiv 2602.23894 ↗First posted 27 Feb 2026cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

Estimating 3D occupancy and motion at the vehicle's surroundings is essential for autonomous driving, enabling situational awareness in dynamic environments. Existing approaches jointly learn geometry and motion but rely on expensive 3D occupancy and flow annotations, velocity labels from bounding boxes, or pretrained optical flow models. We propose a self-supervised method for 3D occupancy flow estimation that eliminates the need for human-produced annotations or external flow supervision. Our method disentangles the scene into separate static and dynamic signed distance fields and learns motion implicitly through temporal aggregation. Additionally, we introduce a strong self-supervised flow cue derived from features' cosine similarities. We demonstrate the efficacy of our 3D occupancy flow method on SemanticKITTI, KITTI-MOT, and nuScenes.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 40 of its 48 references that are in the map. First posted on arXiv 27 Feb 2026.

Builds on

40 of 48 references.

References · 40

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,604
  3. 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 11,979
  4. 2016Feature Pyramid Networks for Object Detectioncited by 5,021
  5. 2019nuScenes: A multimodal dataset for autonomous drivingcited by 4,485
  6. 2019DeepSDF: Learning Continuous Signed Distance Functions for Shape Representationcited by 2,682
  7. 2018Occupancy Networks: Learning 3D Reconstruction in Function Spacecited by 2,037
  8. 2021NeuS: Learning Neural Implicit Surfaces by Volume Rendering for Multi-view Reconstructioncited by 1,701
  9. 2020pixelNeRF: Neural Radiance Fields from One or Few Imagescited by 1,574
  10. 2020D-NeRF: Neural Radiance Fields for Dynamic Scenescited by 1,459
  11. 2022BEVFormer: Learning Bird's-Eye-View Representation from Multi-Camera Images via Spatiotemporal Transformerscited by 1,213
  12. 2015Unsupervised Visual Representation Learning by Context Predictioncited by 1,062
28 more
  1. 2024Grounded SAM: Assembling Open-World Models for Diverse Visual Taskscited by 914
  2. 2023CoTracker: It is Better to Track Togethercited by 516
  3. 2016Semantic Scene Completion from a Single Depth Imagecited by 468
  4. 2023Tri-Perspective View for Vision-Based 3D Semantic Occupancy Predictioncited by 404
  5. 2021MonoScene: Monocular 3D Semantic Scene Completioncited by 357
  6. 2023Spacetime Gaussian Feature Splatting for Real-Time Dynamic View Synthesiscited by 341
  7. 2023SurroundOcc: Multi-Camera 3D Occupancy Prediction for Autonomous Drivingcited by 310
  8. 2022Unifying Flow, Stereo and Depth Estimationcited by 270
  9. 2020Sparse Single Sweep LiDAR Point Cloud Segmentation via Learning Contextual Shape Priors from Scene Completioncited by 222
  10. 2023EmerNeRF: Emergent Spatial-Temporal Scene Decomposition via Self-Supervisioncited by 216
  11. 2023Scene as Occupancycited by 208
  12. 2020LMSCNet: Lightweight Multiscale 3D Semantic Completioncited by 187
  13. 2023SelfOcc: Self-Supervised Vision-Based 3D Occupancy Predictioncited by 133
  14. 2020Anisotropic Convolutional Networks for 3D Semantic Scene Completioncited by 126
  15. 2023RenderOcc: Vision-Centric 3D Occupancy Prediction with 2D Rendering Supervisioncited by 126
  16. 20203D Sketch-aware Semantic Scene Completion via Semi-supervised Structure Priorcited by 124
  17. 2022Vision Transformer for NeRF-Based View Synthesis from a Single Input Imagecited by 104
  18. 2020Semantic Scene Completion using Local Deep Implicit Functions on LiDAR Datacited by 74
  19. 2019Efficient Semantic Scene Completion Network with Spatial Group Convolutioncited by 66
  20. 2023Behind the Scenes: Density Fields for Single View Reconstructioncited by 63
  21. 2023Fully Sparse 3D Occupancy Predictioncited by 61
  22. 2022SceneRF: Self-Supervised Monocular 3D Scene Reconstruction with Radiance Fieldscited by 49
  23. 2018Two Stream 3D Semantic Scene Completioncited by 48
  24. 2023OccNeRF: Advancing 3D Occupancy Prediction in LiDAR-Free Environmentscited by 39
  25. 2024ViewFormer: Exploring Spatiotemporal Modeling for Multi-View 3D Occupancy Perception via View-Guided Transformerscited by 36
  26. 2024Let Occ Flow: Self-Supervised 3D Occupancy Flow Predictioncited by 23
  27. 2024OccFlowNet: Towards Self-supervised Occupancy Estimation via Differentiable Rendering and Occupancy Flowcited by 22
  28. 2024Multi-modal NeRF Self-Supervision for LiDAR Semantic Segmentationcited by 5

Built on by

Not cited yet.