Ethical AI for Biomedical Imaging
summary
An open-source Nime extension implements image-processing nodes (JPEG/PNG input, denoising, equalization, edge/contour extraction) by bridging Java core with Python and OpenCV code, letting users build codeless pipelines and evaluate results. AI nodes include YOLO (detects ~50 categories, returns bounding boxes and masks) and Segment Anything Medical applied to a Kaggle breast-lesion set; workflows train on synthetic data to detect real objects and produce accuracy plots. Nodes are functional for demos and early medical use when an initial region is provided.
AI-written summary · about
chapters (6)
- 0:00Introduction to talk
- 1:33NIME platform for images
- 4:26Image processing pipeline demo
- 6:34AI nodes for segmentation
- 10:53Medical image analysis with SAM
- 13:24Project contributors and resources
related talks
06mTransitioning AI from the lab to real-life applications
31mAI for Everyone: Unleashing the Power of Open Source Tools
33mAI Images: Redefining Quality with Innovation
23mTrusted Agentic AI:Architecting Systems with Reliable Data at Scale
36mApps That See 👀 Practical AI Vision for Developers
29mWhat the AI revolution means for Open Source and our society
44mThe age of applied AI
39mAI is everywhere!