AutoML: No Code No Problem
summary
Google AutoML Vision claims to let non-coders build custom image classifiers; the demo uses the AutoML UI, a ZIP or CSV import, and the AutoML API to auto-add predictions. A poster classifier trains on 16,000 images with useful results after a few hours; baseline is ~100 images per label. Training is free one hour then $20/hour, multi-label drops the confusion matrix, GIFs use the first frame, and evaluation shows real false positives and a usable confusion matrix for up to ten labels.
AI-written summary · about
chapters (6)
- 0:00Start
- 2:40User Interface
- 6:26Tips and Tricks of Auto Ml
- 8:46Multi-Label Classification
- 14:00Image Detector
- 20:59Confusion Matrix
related talks
22mAutoML Vision: the PaaS of Machine Learning
42mUse Machine Learning in your code, without being a ML expert
22mTrain Custom Machine Learning Models with No Data Science Expertise
47mBuilding smarter solutions without machine learning expertise
57mBuilding smarter solutions with Machine Learning, from magic to reality
16mHow to get started with Computer Vision AutoML in Azure
35mComputer vision: British Clay
29mHow to Effortlessly Automate File Recognition with AutoML