mlsanityData to 7 Oct 2026

UNIT: Unsupervised Online Instance Segmentation through Time

arXiv 2409.07887 ↗First posted 12 Sep 2024cs.CVcited by 5

Not on the map. Why?

Abstract (arXiv)

Online object segmentation and tracking in Lidar point clouds enables autonomous agents to understand their surroundings and make safe decisions. Unfortunately, manual annotations for these tasks are prohibitively costly. We tackle this problem with the task of class-agnostic unsupervised online instance segmentation and tracking. To that end, we leverage an instance segmentation backbone and propose a new training recipe that enables the online tracking of objects. Our network is trained on pseudo-labels, eliminating the need for manual annotations. We conduct an evaluation using metrics adapted for temporal instance segmentation. Computing these metrics requires temporally-consistent instance labels. When unavailable, we construct these labels using the available 3D bounding boxes and semantic labels in the dataset. We compare our method against strong baselines and demonstrate its superiority across two different outdoor Lidar datasets.

Key facts. Cited by 5 papers in the map (data through 7 Oct 2026). Builds on 30 of its 40 references that are in the map. First posted on arXiv 12 Sep 2024.

Builds on

30 of 40 references.

References · 30

  1. 2023DINOv2: Learning Robust Visual Features without Supervisioncited by 7,764
  2. 2020End-to-End Object Detection with Transformerscited by 7,384
  3. 2021Masked Autoencoders Are Scalable Vision Learnerscited by 7,339
  4. 2019nuScenes: A multimodal dataset for autonomous drivingcited by 4,485
  5. 2022FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awarenesscited by 3,611
  6. 2021Masked-attention Mask Transformer for Universal Image Segmentationcited by 2,338
  7. 20194D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networkscited by 1,201
  8. 2021Per-Pixel Classification is Not All You Need for Semantic Segmentationcited by 1,148
  9. 2021TrackFormer: Multi-Object Tracking with Transformerscited by 515
  10. 2020PointContrast: Unsupervised Pre-training for 3D Point Cloud Understandingcited by 502
  11. 2021MOTR: End-to-End Multiple-Object Tracking with Transformercited by 379
  12. 2020TransTrack: Multiple Object Tracking with Transformercited by 370
18 more
  1. 2021An End-to-End Transformer Model for 3D Object Detectioncited by 329
  2. 2022Mask3D: Mask Transformer for 3D Semantic Instance Segmentationcited by 299
  3. 2021PandaSet: Advanced Sensor Suite Dataset for Autonomous Drivingcited by 186
  4. 2021Panoptic nuScenes: A Large-Scale Benchmark for LiDAR Panoptic Segmentation and Trackingcited by 165
  5. 2020Train in Germany, Test in The USA: Making 3D Object Detectors Generalizecited by 150
  6. 2022Image-to-Lidar Self-Supervised Distillation for Autonomous Driving Datacited by 126
  7. 2022ALSO: Automotive Lidar Self-supervision by Occupancy estimationcited by 68
  8. 20214D Panoptic LiDAR Segmentationcited by 64
  9. 2020MOPT: Multi-Object Panoptic Trackingcited by 53
  10. 2022Patchwork++: Fast and Robust Ground Segmentation Solving Partial Under-Segmentation Using 3D Point Cloudcited by 45
  11. 2023Three Pillars improving Vision Foundation Model Distillation for Lidarcited by 43
  12. 2023UnScene3D: Unsupervised 3D Instance Segmentation for Indoor Scenescited by 31
  13. 2021Patchwork: Concentric Zone-based Region-wise Ground Segmentation with Ground Likelihood Estimation Using a 3D LiDAR Sensorcited by 28
  14. 2023BEVContrast: Self-Supervision in BEV Space for Automotive Lidar Point Cloudscited by 24
  15. 2023Spatiotemporal Self-supervised Learning for Point Clouds in the Wildcited by 23
  16. 2024LISO: Lidar-only Self-Supervised 3D Object Detectioncited by 22
  17. 2024SeMoLi: What Moves Together Belongs Togethercited by 13
  18. 2024AutoInst: Automatic Instance-Based Segmentation of LiDAR 3D Scanscited by 3

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 3; 2026: 2.

5 papers.

  1. 2025Forging Spatial Intelligence: A Roadmap of Multi-Modal Data Pre-Training for Autonomous Systemscited by 4
  2. 2025Is clustering enough for LiDAR instance segmentation? A state-of-the-art training-free baselinecited by 1
  3. 2026UniLiPs: Unified LiDAR Pseudo-Labeling with Geometry-Grounded Dynamic Scene Decompositioncited by 1
  4. 2025Unsupervised Online 3D Instance Segmentation with Synthetic Sequences and Dynamic Losscited by 0
  5. 2026Emergent 3D Instance Segmentation from Self-Supervised Point Transformerscited by 0