Supervised Fine-tuning in turn Improves Visual Foundation Models
This work addresses scalability issues in fine-grained visual learning for foundation models, offering an incremental improvement over existing methods.
The paper tackles the challenge of enhancing vision foundation models by proposing ViSFT, a two-stage fine-tuning method that improves performance on out-of-domain benchmarks, achieving gains across vision and vision-linguistic tasks with a 4.4B-parameter model trained in under 2 days.
Image-text training like CLIP has dominated the pretraining of vision foundation models in recent years. Subsequent efforts have been made to introduce region-level visual learning into CLIP's pretraining but face scalability challenges due to the lack of large-scale region-level datasets. Drawing inspiration from supervised fine-tuning (SFT) in natural language processing such as instruction tuning, we explore the potential of fine-grained SFT in enhancing the generation of vision foundation models after their pretraining. Thus a two-stage method ViSFT (Vision SFT) is proposed to unleash the fine-grained knowledge of vision foundation models. In ViSFT, the vision foundation model is enhanced by performing visual joint learning on some in-domain tasks and then tested on out-of-domain benchmarks. With updating using ViSFT on 8 V100 GPUs in less than 2 days, a vision transformer with over 4.4B parameters shows improvements across various out-of-domain benchmarks including vision and vision-linguistic scenarios.