# VISA: VLM-Guided Instance Semantic Auditing for 3D Occupancy World Models

arXiv: https://arxiv.org/abs/2606.13460 · HTML: https://mlsanity.com/paper/2606.13460/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in World models for autonomous driving, one of 11 directions in World models.
- Builds on 47 of its 51 references that are in the map.
- First posted on arXiv 11 Jun 2026.

## Placed in

- World models / [World models for autonomous driving](https://mlsanity.com/wm/world-models-for-autonomous-driving/)

## Builds on (most cited first, 30 of 47)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Visual Instruction Tuning](https://mlsanity.com/paper/2304.08485/) (2023)
- [nuScenes: A multimodal dataset for autonomous driving](https://mlsanity.com/paper/1903.11027/) (2019)
- [Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution](https://mlsanity.com/paper/2409.12191/) (2024)
- [Sigmoid Loss for Language Image Pre-Training](https://mlsanity.com/paper/2303.15343/) (2023)
- [Supervised Contrastive Learning](https://mlsanity.com/paper/2004.11362/) (2020)
- [InternVL: Scaling up Vision Foundation Models and Aligning for Generic Visual-Linguistic Tasks](https://mlsanity.com/paper/2312.14238/) (2023)
- [Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond](https://mlsanity.com/paper/2308.12966/) (2023)
- [World Models](https://mlsanity.com/paper/1803.10122/) (2018)
- [Class-Balanced Loss Based on Effective Number of Samples](https://mlsanity.com/paper/1901.05555/) (2019)
- [BEVFormer: Learning Bird's-Eye-View Representation from Multi-Camera Images via Spatiotemporal Transformers](https://mlsanity.com/paper/2203.17270/) (2022)
- [Decoupling Representation and Classifier for Long-Tailed Recognition](https://mlsanity.com/paper/1910.09217/) (2019)
- [Semantic Scene Completion from a Single Depth Image](https://mlsanity.com/paper/1611.08974/) (2016)
- [Tri-Perspective View for Vision-Based 3D Semantic Occupancy Prediction](https://mlsanity.com/paper/2302.07817/) (2023)
- [MonoScene: Monocular 3D Semantic Scene Completion](https://mlsanity.com/paper/2112.00726/) (2021)
- [Occ3D: A Large-Scale 3D Occupancy Prediction Benchmark for Autonomous Driving](https://mlsanity.com/paper/2304.14365/) (2023)
- [SurroundOcc: Multi-Camera 3D Occupancy Prediction for Autonomous Driving](https://mlsanity.com/paper/2303.09551/) (2023)
- [VoxFormer: Sparse Voxel Transformer for Camera-based 3D Semantic Scene Completion](https://mlsanity.com/paper/2302.12251/) (2023)
- [OccFormer: Dual-path Transformer for Vision-based 3D Semantic Occupancy Prediction](https://mlsanity.com/paper/2304.05316/) (2023)
- [MM-LLMs: Recent Advances in MultiModal Large Language Models](https://mlsanity.com/paper/2401.13601/) (2024)
- [OccWorld: Learning a 3D Occupancy World Model for Autonomous Driving](https://mlsanity.com/paper/2311.16038/) (2023)
- [OpenOccupancy: A Large Scale Benchmark for Surrounding Semantic Occupancy Perception](https://mlsanity.com/paper/2303.03991/) (2023)
- [Atlas: End-to-End 3D Scene Reconstruction from Posed Images](https://mlsanity.com/paper/2003.10432/) (2020)
- [PLA: Language-Driven Open-Vocabulary 3D Scene Understanding](https://mlsanity.com/paper/2211.16312/) (2022)
- [FB-OCC: 3D Occupancy Prediction based on Forward-Backward View Transformation](https://mlsanity.com/paper/2307.01492/) (2023)
- [GaussianFormer: Scene as Gaussians for Vision-Based 3D Semantic Occupancy Prediction](https://mlsanity.com/paper/2405.17429/) (2024)
- [SelfOcc: Self-Supervised Vision-Based 3D Occupancy Prediction](https://mlsanity.com/paper/2311.12754/) (2023)
- [RenderOcc: Vision-Centric 3D Occupancy Prediction with 2D Rendering Supervision](https://mlsanity.com/paper/2309.09502/) (2023)
- [PanoOcc: Unified Occupancy Representation for Camera-based 3D Panoptic Segmentation](https://mlsanity.com/paper/2306.10013/) (2023)
- [RegionPLC: Regional Point-Language Contrastive Learning for Open-World 3D Scene Understanding](https://mlsanity.com/paper/2304.00962/) (2023)

## Abstract (arXiv)

Semantic 3D occupancy provides a voxelized world state for autonomous driving and robot decision making, but object and rare-class errors can affect free-space interpretation, collision checking, and temporal state propagation. We show that a common VLM strategy, aligning 3D voxel or object features with crop-caption embeddings, improves text-space similarity without reliably improving closed-set occupancy mIoU. Motivated by this mismatch, we propose VISA, a training-time semantic auditing approach for existing occupancy world models. VISA queries an offline VLM on a representative crop of each physical object instance, obtains a structured audit with class hypotheses, plausible confusions, reliability, attributes, and evidence, and propagates it along the object track. The audit is grounded to matched 3D object voxels and distilled into semantic logits through reliability-weighted taxonomy, attribute-factor, and scene-level audit graph losses, while inference remains unchanged and requires no VLM. On nuScenes, averaged across three runs, VISA improves OccWorld from 19.06 to 20.05 mIoU and GaussianWorld from 21.36 to 21.91 mIoU; on GaussianWorld, object mIoU improves from 18.18 to 19.16 and rare-class mIoU from 15.60 to 16.79. These results suggest that VLMs are better suited to closed-set occupancy as reliability-aware semantic auditors than as generic caption-embedding targets.