# Exploring 2D backbone effects for indoor semantic occupancy prediction

arXiv: https://arxiv.org/abs/2609.17257 · HTML: https://mlsanity.com/paper/2609.17257/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 16 of its 20 references that are in the map.
- First posted on arXiv 15 Sep 2026.

## Builds on (most cited first, 16 of 16)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [DINOv2: Learning Robust Visual Features without Supervision](https://mlsanity.com/paper/2304.07193/) (2023)
- [BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models](https://mlsanity.com/paper/2301.12597/) (2023)
- [Feature Pyramid Networks for Object Detection](https://mlsanity.com/paper/1612.03144/) (2016)
- [4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks](https://mlsanity.com/paper/1904.08755/) (2019)
- [Tri-Perspective View for Vision-Based 3D Semantic Occupancy Prediction](https://mlsanity.com/paper/2302.07817/) (2023)
- [VoxFormer: Sparse Voxel Transformer for Camera-based 3D Semantic Scene Completion](https://mlsanity.com/paper/2302.12251/) (2023)
- [EmbodiedScan: A Holistic Multi-Modal 3D Perception Suite Towards Embodied AI](https://mlsanity.com/paper/2312.16170/) (2023)
- [SelfOcc: Self-Supervised Vision-Based 3D Occupancy Prediction](https://mlsanity.com/paper/2311.12754/) (2023)
- [OccVLA: Vision-Language-Action Model with Implicit 3D Occupancy Supervision](https://mlsanity.com/paper/2509.05578/) (2025)
- [OccTENS: 3D Occupancy World Model via Temporal Next-Scale Prediction](https://mlsanity.com/paper/2509.03887/) (2025)
- [Semantic Causality-Aware Vision-Based 3D Occupancy Prediction](https://mlsanity.com/paper/2509.08388/) (2025)
- [ShelfOcc: Native 3D Supervision beyond LiDAR for Vision-Based Occupancy Estimation](https://mlsanity.com/paper/2511.15396/) (2025)
- [OccAny: Generalized Unconstrained Urban 3D Occupancy](https://mlsanity.com/paper/2603.23502/) (2026)
- [SliceSemOcc: Vertical Slice Based Multimodal 3D Semantic Occupancy Representation](https://mlsanity.com/paper/2509.03999/) (2025)

## Abstract (arXiv)

Semantic occupancy prediction gives an embodied agent a voxel-level account of where space is free, occupied, and semantically meaningful. In RGB-D pipelines such as EmbodiedScan, the image encoder is often left as a default module, even though its features are the visual evidence later sampled into the 3D grid. We study this design choice directly. A central finding is that changing the 2D backbone improves occupancy accuracy more than several carefully designed occupancy architectures or modules. We keep the main RGB-D projection, depth branch, and occupancy head fixed, and replace only the image backbone. The compared encoders are CLIP-ResNet, CLIP-ViT, BLIP2, and DINOv2. Under the controlled setting, the measured mIoU changes substantially: DINOv2 obtains 30.55%, BLIP2 obtains 29.49%, CLIP-ViT obtains 24.33%, and CLIP-ResNet obtains 17.41%. The stronger encoders also exceed the original EmbodiedScan ResNet-50 baseline without modifying the downstream 3D fusion pipeline. Class-level results give a more detailed picture: DINOv2 is stronger on many layout and structural categories, whereas BLIP2 remains close on several object-centered classes. CLIP-ViT improves clearly over CLIP-ResNet, showing that the way CLIP features are exposed as dense tokens matters for voxel lifting. These results indicate that the image backbone is not a secondary engineering detail in embodied semantic occupancy, but a major source of variation in the final 3D prediction.