Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2503.20322 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917247257149440 |
|---|---|
| author | Ai, Hao Wang, Kunyi Wang, Zezhou Lu, Hao Tian, Jin Luo, Yaxin Xing, Peng Huang, Jen-Yuan Li, Huaxia luo, Gen |
| author_facet | Ai, Hao Wang, Kunyi Wang, Zezhou Lu, Hao Tian, Jin Luo, Yaxin Xing, Peng Huang, Jen-Yuan Li, Huaxia luo, Gen |
| contents | Multimodal large language models (MLLMs) have demonstrated impressive performance in various vision-language (VL) tasks, but their expensive computations still limit the real-world application. To address this issue, recent efforts aim to compress the visual features to save the computational costs of MLLMs. However, direct visual compression methods, e.g. efficient projectors, inevitably destroy the visual semantics in MLLM, especially in difficult samples. To overcome this shortcoming, we propose a novel dynamic pyramid network (DPN) for efficient MLLMs. Specifically, DPN formulates MLLM as a hierarchical structure where visual features are gradually compressed with increasing depth. In this case, even with a high compression ratio, fine-grained visual information can still be perceived in shallow layers. To maximize the benefit of DPN, we further propose an innovative Dynamic Pooling Experts (DPE) that can dynamically choose the optimal visual compression rate according to input features. With this design, harder samples will be assigned larger computations, thus preserving the model performance. To validate our approach, we conduct extensive experiments on two popular MLLMs and ten benchmarks. Experimental results show that DPN can save up to 56% average FLOPs on LLaVA while further achieving +0.74% performance gains. Besides, the generalization ability of DPN is also validated on the existing high-resolution MLLM called LLaVA-HR. The source code will be released at https://github.com/aihao2000/DPN-LLaVA. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_20322 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Dynamic Pyramid Network for Efficient Multimodal Large Language Model Ai, Hao Wang, Kunyi Wang, Zezhou Lu, Hao Tian, Jin Luo, Yaxin Xing, Peng Huang, Jen-Yuan Li, Huaxia luo, Gen Computer Vision and Pattern Recognition Multimodal large language models (MLLMs) have demonstrated impressive performance in various vision-language (VL) tasks, but their expensive computations still limit the real-world application. To address this issue, recent efforts aim to compress the visual features to save the computational costs of MLLMs. However, direct visual compression methods, e.g. efficient projectors, inevitably destroy the visual semantics in MLLM, especially in difficult samples. To overcome this shortcoming, we propose a novel dynamic pyramid network (DPN) for efficient MLLMs. Specifically, DPN formulates MLLM as a hierarchical structure where visual features are gradually compressed with increasing depth. In this case, even with a high compression ratio, fine-grained visual information can still be perceived in shallow layers. To maximize the benefit of DPN, we further propose an innovative Dynamic Pooling Experts (DPE) that can dynamically choose the optimal visual compression rate according to input features. With this design, harder samples will be assigned larger computations, thus preserving the model performance. To validate our approach, we conduct extensive experiments on two popular MLLMs and ten benchmarks. Experimental results show that DPN can save up to 56% average FLOPs on LLaVA while further achieving +0.74% performance gains. Besides, the generalization ability of DPN is also validated on the existing high-resolution MLLM called LLaVA-HR. The source code will be released at https://github.com/aihao2000/DPN-LLaVA. |
| title | Dynamic Pyramid Network for Efficient Multimodal Large Language Model |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2503.20322 |