SAMJ: Fast Image Annotation on ImageJ/Fiji via Segment Anything Model
🤖 AI Summary
Manual annotation of biomedical image masks is inefficient and severely hinders AI model training and deployment. To address this, we present the first lightweight ImageJ/Fiji plugin seamlessly integrating the Segment Anything Model (SAM), enabling one-click installation and zero-code interactive annotation. Our method introduces CPU/GPU-adaptive inference scheduling, optimized inference for large-scale scientific images, and intuitive point-and-box prompting interfaces—collectively enabling sub-second mask generation. On mainstream research-grade hardware, annotation throughput improves by over 10×, supporting efficient processing of terabyte-scale microscopy datasets. The plugin has been officially integrated into the Fiji update site and is now widely adopted as a community-standard annotation tool. This work significantly lowers both the technical barrier and hardware requirements for AI-driven biomedical image analysis.
Application Category
📝 Abstract
Mask annotation remains a significant bottleneck in AI-driven biomedical image analysis due to its labor-intensive nature. To address this challenge, we introduce SAMJ, a user-friendly ImageJ/Fiji plugin leveraging the Segment Anything Model (SAM). SAMJ enables seamless, interactive annotations with one-click installation on standard computers. Designed for real-time object delineation in large scientific images, SAMJ is an easy-to-use solution that simplifies and accelerates the creation of labeled image datasets.
Problem
Research questions and friction points this paper is trying to address.
Labor-intensive mask annotation in biomedical image analysis
Need for user-friendly interactive annotation tools
Challenges in real-time object delineation for large images
Innovation
Methods, ideas, or system contributions that make the work stand out.
ImageJ plugin for fast annotation
Leverages Segment Anything Model
One-click installation for ease
🔎 Similar Papers
No similar papers found.
