Saved in:
Bibliographic Details
Main Authors: Patil, Kunal Dasharath, Palani, Gowthamaan, Krishnamurthi, Ganapathy
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2501.16740
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929689426132992
author Patil, Kunal Dasharath
Palani, Gowthamaan
Krishnamurthi, Ganapathy
author_facet Patil, Kunal Dasharath
Palani, Gowthamaan
Krishnamurthi, Ganapathy
contents The Segment Anything Model (SAM) has set a new standard in interactive image segmentation, offering robust performance across various tasks. However, its significant computational requirements limit its deployment in real-time or resource-constrained environments. To address these challenges, we propose a novel knowledge distillation approach, KD SAM, which incorporates both encoder and decoder optimization through a combination of Mean Squared Error (MSE) and Perceptual Loss. This dual-loss framework captures structural and semantic features, enabling the student model to maintain high segmentation accuracy while reducing computational complexity. Based on the model evaluation on datasets, including Kvasir-SEG, ISIC 2017, Fetal Head Ultrasound, and Breast Ultrasound, we demonstrate that KD SAM achieves comparable or superior performance to the baseline models, with significantly fewer parameters. KD SAM effectively balances segmentation accuracy and computational efficiency, making it well-suited for real-time medical image segmentation applications in resource-constrained environments.
format Preprint
id arxiv_https___arxiv_org_abs_2501_16740
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Efficient Knowledge Distillation of SAM for Medical Image Segmentation
Patil, Kunal Dasharath
Palani, Gowthamaan
Krishnamurthi, Ganapathy
Image and Video Processing
Artificial Intelligence
Computer Vision and Pattern Recognition
The Segment Anything Model (SAM) has set a new standard in interactive image segmentation, offering robust performance across various tasks. However, its significant computational requirements limit its deployment in real-time or resource-constrained environments. To address these challenges, we propose a novel knowledge distillation approach, KD SAM, which incorporates both encoder and decoder optimization through a combination of Mean Squared Error (MSE) and Perceptual Loss. This dual-loss framework captures structural and semantic features, enabling the student model to maintain high segmentation accuracy while reducing computational complexity. Based on the model evaluation on datasets, including Kvasir-SEG, ISIC 2017, Fetal Head Ultrasound, and Breast Ultrasound, we demonstrate that KD SAM achieves comparable or superior performance to the baseline models, with significantly fewer parameters. KD SAM effectively balances segmentation accuracy and computational efficiency, making it well-suited for real-time medical image segmentation applications in resource-constrained environments.
title Efficient Knowledge Distillation of SAM for Medical Image Segmentation
topic Image and Video Processing
Artificial Intelligence
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2501.16740