OneLLM: One Framework to Align All Modalities with Language
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2023
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913641771565056 |
|---|---|
| author | Han, Jiaming Gong, Kaixiong Zhang, Yiyuan Wang, Jiaqi Zhang, Kaipeng Lin, Dahua Qiao, Yu Gao, Peng Yue, Xiangyu |
| author_facet | Han, Jiaming Gong, Kaixiong Zhang, Yiyuan Wang, Jiaqi Zhang, Kaipeng Lin, Dahua Qiao, Yu Gao, Peng Yue, Xiangyu |
| contents | Multimodal large language models (MLLMs) have gained significant attention due to their strong multimodal understanding capability. However, existing works rely heavily on modality-specific encoders, which usually differ in architecture and are limited to common modalities. In this paper, we present OneLLM, an MLLM that aligns eight modalities to language using a unified framework. We achieve this through a unified multimodal encoder and a progressive multimodal alignment pipeline. In detail, we first train an image projection module to connect a vision encoder with LLM. Then, we build a universal projection module (UPM) by mixing multiple image projection modules and dynamic routing. Finally, we progressively align more modalities to LLM with the UPM. To fully leverage the potential of OneLLM in following instructions, we also curated a comprehensive multimodal instruction dataset, including 2M items from image, audio, video, point cloud, depth/normal map, IMU and fMRI brain activity. OneLLM is evaluated on 25 diverse benchmarks, encompassing tasks such as multimodal captioning, question answering and reasoning, where it delivers excellent performance. Code, data, model and online demo are available at https://github.com/csuhan/OneLLM |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2312_03700 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | OneLLM: One Framework to Align All Modalities with Language Han, Jiaming Gong, Kaixiong Zhang, Yiyuan Wang, Jiaqi Zhang, Kaipeng Lin, Dahua Qiao, Yu Gao, Peng Yue, Xiangyu Computer Vision and Pattern Recognition Artificial Intelligence Computation and Language Machine Learning Multimedia Multimodal large language models (MLLMs) have gained significant attention due to their strong multimodal understanding capability. However, existing works rely heavily on modality-specific encoders, which usually differ in architecture and are limited to common modalities. In this paper, we present OneLLM, an MLLM that aligns eight modalities to language using a unified framework. We achieve this through a unified multimodal encoder and a progressive multimodal alignment pipeline. In detail, we first train an image projection module to connect a vision encoder with LLM. Then, we build a universal projection module (UPM) by mixing multiple image projection modules and dynamic routing. Finally, we progressively align more modalities to LLM with the UPM. To fully leverage the potential of OneLLM in following instructions, we also curated a comprehensive multimodal instruction dataset, including 2M items from image, audio, video, point cloud, depth/normal map, IMU and fMRI brain activity. OneLLM is evaluated on 25 diverse benchmarks, encompassing tasks such as multimodal captioning, question answering and reasoning, where it delivers excellent performance. Code, data, model and online demo are available at https://github.com/csuhan/OneLLM |
| title | OneLLM: One Framework to Align All Modalities with Language |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Computation and Language Machine Learning Multimedia |
| url | https://arxiv.org/abs/2312.03700 |