# SPACE-CLIPv2: Decoding Local Geometry from Frozen CLIP for Monocular Depth Estimation

arXiv: https://arxiv.org/abs/2610.05029 · HTML: https://mlsanity.com/paper/2610.05029/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 22 of its 30 references that are in the map.
- First posted on arXiv 4 Oct 2026.

## Placed in

- Multimodal LLMs / vision-language models

## Builds on (most cited first, 22 of 22)

- [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)
- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://mlsanity.com/paper/2103.14030/) (2021)
- [Vision Transformers for Dense Prediction](https://mlsanity.com/paper/2103.13413/) (2021)
- [Deformable Convolutional Networks](https://mlsanity.com/paper/1703.06211/) (2017)
- [Depth Anything: Unleashing the Power of Large-Scale Unlabeled Data](https://mlsanity.com/paper/2401.10891/) (2024)
- [Depth Map Prediction from a Single Image using a Multi-Scale Deep Network](https://mlsanity.com/paper/1406.2283/) (2014)
- [ZoeDepth: Zero-shot Transfer by Combining Relative and Metric Depth](https://mlsanity.com/paper/2302.12288/) (2023)
- [AdaBins: Depth Estimation using Adaptive Bins](https://mlsanity.com/paper/2011.14141/) (2020)
- [DenseCLIP: Language-Guided Dense Prediction with Context-Aware Prompting](https://mlsanity.com/paper/2112.01518/) (2021)
- [Metric3D: Towards Zero-shot Metric 3D Prediction from A Single Image](https://mlsanity.com/paper/2307.10984/) (2023)
- [DIODE: A Dense Indoor and Outdoor DEpth Dataset](https://mlsanity.com/paper/1908.00463/) (2019)
- [BinsFormer: Revisiting Adaptive Bins for Monocular Depth Estimation](https://mlsanity.com/paper/2204.00987/) (2022)
- [Evaluation of CNN-based Single-Image Depth Estimation Methods](https://mlsanity.com/paper/1805.01328/) (2018)
- [Can Language Understand Depth?](https://mlsanity.com/paper/2207.01077/) (2022)
- [Learning to Adapt CLIP for Few-Shot Monocular Depth Estimation](https://mlsanity.com/paper/2311.01034/) (2023)
- [PrimeDepth: Efficient Monocular Depth Estimation with a Stable Diffusion Preimage](https://mlsanity.com/paper/2409.09144/) (2024)
- [Structure-Centric Robust Monocular Depth Estimation via Knowledge Distillation](https://mlsanity.com/paper/2410.06982/) (2024)
- [SPACE-CLIP: Spatial Perception via Adaptive CLIP Embeddings for Monocular Depth Estimation](https://mlsanity.com/paper/2601.17657/) (2026)
- [PureCLIP-Depth: Prompt-Free and Decoder-Free Monocular Depth Estimation within CLIP Embedding Space](https://mlsanity.com/paper/2603.16238/) (2026)

## Abstract (arXiv)

Vision-language foundation models such as CLIP provide strong semantic representations, but their patch tokens are not directly optimized for dense metric geometry. SPACE-CLIP showed that frozen CLIP features can support monocular depth estimation through layer-group feature fusion, yet it leaves open how neighboring CLIP tokens should be combined to recover fine local structure. We present SPACE-CLIPv2, a frozen-backbone depth decoder that aggregates fixed local neighborhoods in CLIP token space. At selected decoder stages, the model samples a fixed token stencil, predicts aggregation weights, and injects the resulting response through a gated residual update. A token-space high-pass branch further preserves shallow local contrast. On NYU Depth V2, SPACE-CLIPv2 improves over a matched SPACE-CLIP baseline, while five-seed experiments consistently favor fixed over learned-offset sampling. Zero-shot iBims-1 evaluation further improves boundary and planar-geometry measures. These results support constrained local token aggregation as a practical mechanism for decoding geometry from frozen CLIP representations.