mlsanityData to 14 Oct 2026

LOSC: LiDAR Open-voc Segmentation Consolidator

arXiv 2507.07605 ↗First posted 10 Jul 2025cs.CVcited by 5

On the map: Multimodal LLMs / vision-language models → Embodied and 3D vision-language models

Abstract (arXiv)

We study the use of image-based Vision-Language Models (VLMs) for open-vocabulary segmentation of lidar scans in driving settings. Classically, image semantics can be back-projected onto 3D point clouds. Yet, resulting point labels are noisy and sparse. We consolidate these labels to enforce both spatio-temporal consistency and robustness to image-level augmentations. We then train a 3D network based on these refined labels. This simple method, called LOSC, outperforms the SOTA of zero-shot open-vocabulary semantic and panoptic segmentation on both nuScenes and SemanticKITTI, with significant margins. Code is available at https://github.com/valeoai/LOSC.

Key facts. Cited by 5 papers in the map (data through 14 Oct 2026). Placed in Embodied and 3D vision-language models, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 33 of its 48 references that are in the map. First posted on arXiv 10 Jul 2025.

Builds on

33 of 48 references.

Closest ancestors · 2

Same direction, newest first

  1. 2024When LLMs step into the 3D World: A Survey and Meta-Analysis of 3D Tasks via Multi-modal Large Language Models16 May
  2. 2023Towards Label-free Scene Understanding by Vision Foundation Models6 Jun

Other references · 31

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  2. 2023Segment Anythingcited by 8,234
  3. 2019nuScenes: A multimodal dataset for autonomous drivingcited by 4,523
  4. 2022LAION-5B: An open large-scale dataset for training next generation image-text modelscited by 4,152
  5. 2021Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervisioncited by 3,864
  6. 2024Grounded SAM: Assembling Open-World Models for Diverse Visual Taskscited by 929
  7. 2023Vision-Language Models for Vision Tasks: A Surveycited by 839
  8. 2022Language-driven Semantic Segmentationcited by 717
  9. 2021RegionCLIP: Region-based Language-Image Pretrainingcited by 616
  10. 2023Point Transformer V3: Simpler, Faster, Strongercited by 589
  11. 2021Extract Free Dense Labels from CLIPcited by 583
  12. 2018Albumentations: fast and flexible image augmentationscited by 560
19 more
  1. 2022OpenScene: 3D Scene Understanding with Open Vocabulariescited by 559
  2. 2023Segment Everything Everywhere All at Oncecited by 548
  3. 2021Scaling Open-Vocabulary Image Segmentation with Image-Level Labelscited by 473
  4. 2023A Simple Framework for Open-Vocabulary Segmentation and Detectioncited by 198
  5. 2023CLIP2Scene: Towards Label-efficient 3D Scene Understanding by CLIPcited by 181
  6. 2023Towards Open Vocabulary Learning: A Surveycited by 166
  7. 2022Open-Vocabulary Universal Image Segmentation with MaskCLIPcited by 132
  8. 2023Using a Waffle Iron for Automotive Point Cloud Semantic Segmentationcited by 53
  9. 2024Better Call SAL: Towards Learning to Segment Anything in Lidarcited by 47
  10. 2023Three Pillars improving Vision Foundation Model Distillation for Lidarcited by 44
  11. 2023CLIP-DIY: CLIP Dense Inference Yields Open-Vocabulary Semantic Segmentation For-Freecited by 39
  12. 2024Image Segmentation in Foundation Model Era: A Surveycited by 35
  13. 2024Open Vocabulary 3D Scene Understanding via Geometry Guided Self-Distillationcited by 17
  14. 2025Vision-Language Model for Object Detection and Segmentation: A Review and Evaluationcited by 10
  15. 2025LeAP: Consistent multi-domain 3D labeling using Foundation Modelscited by 8
  16. 2024AdaCo: Overcoming Visual Foundation Model Noise in 3D Semantic Segmentation via Adaptive Label Correctioncited by 6
  17. 2025Zero-Shot 4D Lidar Panoptic Segmentationcited by 6
  18. 20243D Annotation-Free Learning by Distilling 2D Open-Vocabulary Segmentation Models for Autonomous Drivingcited by 5
  19. 2025Image Segmentation with Large Language Models: A Survey with Perspectives for Intelligent Transportation Systemscited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 5.

5 papers.

  1. 2026OccAny: Generalized Unconstrained Urban 3D Occupancycited by 6
  2. 2026Depth-Only Open-Vocabulary 3D Semantic Segmentation For Privacy-Preserving Robotic Applicationscited by 1
  3. 2026Annotation Free Spacecraft Detection and Segmentation using Vision Language Modelscited by 0
  4. 2026IGLOSS: Image Generation for Lidar Open-vocabulary Semantic Segmentationcited by 0
  5. 2026Lang3DSeg: Annotation-Free Open-Vocabulary 3D Segmentation with Point Transformerscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025NUC-Net: Non-uniform Cylindrical Partition Network for Efficient LiDAR Semantic Segmentation
  2. 2026Bridging the Modality Gap in Roadside LiDAR: A Training-Free Vision-Language Model Framework for Vehicle Classification
  3. 2024ELiTe: Efficient Image-to-LiDAR Knowledge Transfer for Semantic Segmentation
  4. 2026Training-Free Open-Vocabulary 3D Point-Cloud Segmentation on the Generalized Few-Shot Benchmark
  5. 2025Improving Multimodal Distillation for 3D Semantic Segmentation under Domain Shift
  6. 2024Global-Local Collaborative Inference with LLM for Lidar-Based Open-Vocabulary Detection
  7. 2024UniMix: Towards Domain Adaptive and Generalizable LiDAR Semantic Segmentation in Adverse Weather
  8. 2024DiffSSC: Semantic LiDAR Scan Completion using Denoising Diffusion Probabilistic Models