Generalized SAM: Efficient Fine-Tuning of SAM for Variable Input Image Sizes
This work addresses the computational inefficiency and information loss in fine-tuning SAM for image segmentation, which is incremental as it builds on existing SAM fine-tuning methods.
The paper tackles the problem of fine-tuning the Segment Anything Model (SAM) for variable input image sizes, which reduces computational costs and avoids information loss from fixed aspect ratios, resulting in GSAM achieving comparable or higher accuracy with more efficient training.
There has been a lot of recent research on improving the efficiency of fine-tuning foundation models. In this paper, we propose a novel efficient fine-tuning method that allows the input image size of Segment Anything Model (SAM) to be variable. SAM is a powerful foundational model for image segmentation trained on huge datasets, but it requires fine-tuning to recognize arbitrary classes. The input image size of SAM is fixed at 1024 x 1024, resulting in substantial computational demands during training. Furthermore, the fixed input image size may result in the loss of image information, e.g. due to fixed aspect ratios. To address this problem, we propose Generalized SAM (GSAM). Different from the previous methods, GSAM is the first to apply random cropping during training with SAM, thereby significantly reducing the computational cost of training. Experiments on datasets of various types and various pixel counts have shown that GSAM can train more efficiently than SAM and other fine-tuning methods for SAM, achieving comparable or higher accuracy.