TinyGiantVLM: A Lightweight Vision-Language Architecture for Spatial Reasoning under Resource Constraints
Fuente:
arXiv
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909751411998720 |
|---|---|
| author | Ly, Vinh-Thuan Truong, Hoang M. Nguyen, Xuan-Huong |
| author_facet | Ly, Vinh-Thuan Truong, Hoang M. Nguyen, Xuan-Huong |
| contents | Reasoning about fine-grained spatial relationships in warehouse-scale environments poses a significant challenge for existing vision-language models (VLMs), which often struggle to comprehend 3D layouts, object arrangements, and multimodal cues in real-world industrial settings. In this paper, we present TinyGiantVLM, a lightweight and modular two-stage framework designed for physical spatial reasoning, distinguishing itself from traditional geographic reasoning in complex logistics scenes. Our approach encodes both global and region-level features from RGB and depth modalities using pretrained visual backbones. To effectively handle the complexity of high-modality inputs and diverse question types, we incorporate a Mixture-of-Experts (MoE) fusion module, which dynamically combines spatial representations to support downstream reasoning tasks and improve convergence. Training is conducted in a two-phase strategy: the first phase focuses on generating free-form answers to enhance spatial reasoning ability, while the second phase uses normalized answers for evaluation. Evaluated on Track 3 of the AI City Challenge 2025, our 64M-parameter base model achieved 5th place on the leaderboard with a score of 66.8861, demonstrating strong performance in bridging visual perception and spatial understanding in industrial environments. We further present an 80M-parameter variant with expanded MoE capacity, which demonstrates improved performance on spatial reasoning tasks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2508_17595 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | TinyGiantVLM: A Lightweight Vision-Language Architecture for Spatial Reasoning under Resource Constraints Ly, Vinh-Thuan Truong, Hoang M. Nguyen, Xuan-Huong Computer Vision and Pattern Recognition Reasoning about fine-grained spatial relationships in warehouse-scale environments poses a significant challenge for existing vision-language models (VLMs), which often struggle to comprehend 3D layouts, object arrangements, and multimodal cues in real-world industrial settings. In this paper, we present TinyGiantVLM, a lightweight and modular two-stage framework designed for physical spatial reasoning, distinguishing itself from traditional geographic reasoning in complex logistics scenes. Our approach encodes both global and region-level features from RGB and depth modalities using pretrained visual backbones. To effectively handle the complexity of high-modality inputs and diverse question types, we incorporate a Mixture-of-Experts (MoE) fusion module, which dynamically combines spatial representations to support downstream reasoning tasks and improve convergence. Training is conducted in a two-phase strategy: the first phase focuses on generating free-form answers to enhance spatial reasoning ability, while the second phase uses normalized answers for evaluation. Evaluated on Track 3 of the AI City Challenge 2025, our 64M-parameter base model achieved 5th place on the leaderboard with a score of 66.8861, demonstrating strong performance in bridging visual perception and spatial understanding in industrial environments. We further present an 80M-parameter variant with expanded MoE capacity, which demonstrates improved performance on spatial reasoning tasks. |
| title | TinyGiantVLM: A Lightweight Vision-Language Architecture for Spatial Reasoning under Resource Constraints |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2508.17595 |