ChatRex: Taming Multimodal LLM for Joint Perception and Understanding
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913729663205376 |
|---|---|
| author | Jiang, Qing Luo, Gen Yang, Yuqin Xiong, Yuda Chen, Yihao Zeng, Zhaoyang Ren, Tianhe Zhang, Lei |
| author_facet | Jiang, Qing Luo, Gen Yang, Yuqin Xiong, Yuda Chen, Yihao Zeng, Zhaoyang Ren, Tianhe Zhang, Lei |
| contents | Perception and understanding are two pillars of computer vision. While multimodal large language models (MLLM) have demonstrated remarkable visual understanding capabilities, they arguably lack accurate perception abilities, e.g. the stage-of-the-art model Qwen2-VL only achieves a 43.9 recall rate on the COCO dataset, limiting many tasks requiring the combination of perception and understanding. In this work, we aim to bridge this perception gap from both model designing and data development perspectives. We first introduce ChatRex, an MLLM with a decoupled perception design. Instead of having the LLM directly predict box coordinates, we feed the output boxes from a universal proposal network into the LLM, allowing it to output the corresponding box indices to represent its detection results, turning the regression task into a retrieval-based task that LLM handles more proficiently. From the data perspective, we build a fully automated data engine and construct the Rexverse-2M dataset which possesses multiple granularities to support the joint training of perception and understanding. After a three-stage training approach, ChatRex demonstrates strong perception and understanding performance, and the combination of these two capabilities also unlocks many attractive applications, demonstrating their complementary roles in MLLM. Code is available at https://github.com/IDEA-Research/ChatRex. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2411_18363 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | ChatRex: Taming Multimodal LLM for Joint Perception and Understanding Jiang, Qing Luo, Gen Yang, Yuqin Xiong, Yuda Chen, Yihao Zeng, Zhaoyang Ren, Tianhe Zhang, Lei Computer Vision and Pattern Recognition Perception and understanding are two pillars of computer vision. While multimodal large language models (MLLM) have demonstrated remarkable visual understanding capabilities, they arguably lack accurate perception abilities, e.g. the stage-of-the-art model Qwen2-VL only achieves a 43.9 recall rate on the COCO dataset, limiting many tasks requiring the combination of perception and understanding. In this work, we aim to bridge this perception gap from both model designing and data development perspectives. We first introduce ChatRex, an MLLM with a decoupled perception design. Instead of having the LLM directly predict box coordinates, we feed the output boxes from a universal proposal network into the LLM, allowing it to output the corresponding box indices to represent its detection results, turning the regression task into a retrieval-based task that LLM handles more proficiently. From the data perspective, we build a fully automated data engine and construct the Rexverse-2M dataset which possesses multiple granularities to support the joint training of perception and understanding. After a three-stage training approach, ChatRex demonstrates strong perception and understanding performance, and the combination of these two capabilities also unlocks many attractive applications, demonstrating their complementary roles in MLLM. Code is available at https://github.com/IDEA-Research/ChatRex. |
| title | ChatRex: Taming Multimodal LLM for Joint Perception and Understanding |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2411.18363 |