File size: 2,051 Bytes
4f95937
 
 
c8e38a1
 
 
4f95937
63234c4
c8e38a1
 
 
 
 
 
 
 
 
 
 
 
 
 
 
63234c4
c8e38a1
63234c4
c8e38a1
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
---
language:
- en
license: apache-2.0
library_name: transformers
pipeline_tag: robotics
---

# FantasyVLN: Unified Multimodal Chain-of-Thought Reasoning for Vision-Language Navigation

**FantasyVLN** is a unified multimodal Chain-of-Thought (CoT) reasoning framework that enables efficient and precise navigation based on natural language instructions and visual observations.

- **Paper:** [FantasyVLN: Unified Multimodal Chain-of-Thought Reasoning for Vision-Language Navigation](https://huggingface.co/papers/2601.13976)
- **Project Page:** [https://fantasy-amap.github.io/fantasy-vln/](https://fantasy-amap.github.io/fantasy-vln/)
- **Code:** [https://github.com/Fantasy-AMAP/fantasy-vln](https://github.com/Fantasy-AMAP/fantasy-vln)

## Introduction

Achieving human-level performance in Vision-and-Language Navigation (VLN) requires an embodied agent to jointly understand multimodal instructions and visual-spatial context while reasoning over long action sequences. 

FantasyVLN combines the benefits of textual, visual, and multimodal CoT reasoning by constructing a unified representation space across these reasoning modes. To enable efficient reasoning, we align these CoT reasoning modes with non-CoT reasoning during training, while using only non-CoT reasoning at test time. Notably, we perform visual CoT in the latent space of a [VAR](https://github.com/FoundationVision/VAR) model, where only low-scale latent representations are predicted. Compared to traditional pixel-level visual CoT methods, our approach significantly improves both training and inference efficiency, reducing inference latency by an order of magnitude compared to explicit CoT methods.

## Citation

If you find this work helpful, please consider citing:

```bibtex
@article{zuo2025fantasyvln,
  title={FantasyVLN: Unified Multimodal Chain-of-Thought Reasoning for Vision-Language Navigation},
  author={Zuo, Jing and Mu, Lingzhou and Jiang, Fan and Ma, Chengcheng and Xu, Mu and Qi, Yonggang},
  journal={arXiv preprint arXiv:2601.13976},
  year={2025}
}
```