JavaScript beats cancer

on YouTube ↗37m

" summary

Claims that JavaScript can run a clinically useful melanoma detector using TensorFlow.js with Keras-trained models served via TensorFlow Serving/Kubeflow, deployed in a React Native mobile app. Uses ISIC research data (from 53 historical images to 26,000 images), extracts and automates about 15 patterns (ABCDE-style features) with preprocessing, segmentation and a points-based scoring threshold; image processing and retraining pipeline work but heavier training and some preprocessing still rely on Python.

AI-written summary · about

" chapters (11)

1. 0:00Start
2. 2:57Agenda
3. 3:25The idea
4. 6:19PoC architecture overview
5. 7:35JavaScript vs other languages
6. 9:32In production
7. 18:34Skin cancer
8. 21:30Scoring
9. 25:30Image processing
10. 33:01Example
11. 33:55Summary

" related talks

• 32mWe Also Can Do It! Machine Learning in Javascript!Eliran Eliassy · DevDays Europe 2019
• 22mKeynote talk - The future of machine learning & JSSimona Cotin · jsPoland 2018
• 15mThe future of machine learning using JavaScriptDoni Rubiagatra · CityJS Singapore 2023
• 15mMachine learning for next gen web apps with TensorFlow.jsJason Mayes · Google I/O 2021
• 51mJoin the RevolutionGant Laborde · JSDay 2019
• 46mArtificial Intelligence with JavaScriptJavier Luraschi · XtremeJS 2021
• 34mThe Future of JavaScript and AIAsim Hussain · JSConf.Asia 2019
• 11mAI-powered JavaScript applicationsAsim Hussain · The Lead Developer London 2019