Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2604.04579 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917387454906368 |
|---|---|
| author | Trinh, Quoc-Huy Abdullahi, Mustapha Zhao, Bo Jha, Debesh |
| author_facet | Trinh, Quoc-Huy Abdullahi, Mustapha Zhao, Bo Jha, Debesh |
| contents | Recent advances in multimodal large language models (MLLMs) have enabled impressive progress in vision-language understanding, yet their high computational cost limits deployment in resource-constrained scenarios such as personal assistants, document understanding, and smart cameras. Most existing methods rely on Transformer-based cross-attention, whose quadratic complexity hinders efficiency. Moreover, small vision-language models often struggle to precisely capture fine-grained, task-relevant visual regions, leading to degraded performance on fine-grained reasoning tasks that limit their effectiveness in the real world. To address these issues, we introduce Firebolt-VL, an efficient vision-language model that replaces the Transformer-based decoder with a Liquid Foundation Model (LFM) decoder. To further enhance visual grounding, we propose a Token-Grid Correlation Module, which computes lightweight correlations between text tokens and image patches and modulates via the state-space model with FiLM conditioning. This enables the model to selectively emphasize visual regions relevant to the textual prompt while maintaining linear-time inference. Experimental results across multiple benchmarks demonstrate that Firebolt-VL achieves accurate, fine-grained understanding with significantly improved efficiency. Our model and code are available at: https://fireboltvl.github.io |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_04579 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Firebolt-VL: Efficient Vision-Language Understanding with Cross-Modality Modulation Trinh, Quoc-Huy Abdullahi, Mustapha Zhao, Bo Jha, Debesh Computer Vision and Pattern Recognition Recent advances in multimodal large language models (MLLMs) have enabled impressive progress in vision-language understanding, yet their high computational cost limits deployment in resource-constrained scenarios such as personal assistants, document understanding, and smart cameras. Most existing methods rely on Transformer-based cross-attention, whose quadratic complexity hinders efficiency. Moreover, small vision-language models often struggle to precisely capture fine-grained, task-relevant visual regions, leading to degraded performance on fine-grained reasoning tasks that limit their effectiveness in the real world. To address these issues, we introduce Firebolt-VL, an efficient vision-language model that replaces the Transformer-based decoder with a Liquid Foundation Model (LFM) decoder. To further enhance visual grounding, we propose a Token-Grid Correlation Module, which computes lightweight correlations between text tokens and image patches and modulates via the state-space model with FiLM conditioning. This enables the model to selectively emphasize visual regions relevant to the textual prompt while maintaining linear-time inference. Experimental results across multiple benchmarks demonstrate that Firebolt-VL achieves accurate, fine-grained understanding with significantly improved efficiency. Our model and code are available at: https://fireboltvl.github.io |
| title | Firebolt-VL: Efficient Vision-Language Understanding with Cross-Modality Modulation |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2604.04579 |