Amazon Lookout for Vision
summary
Service automates defect detection for manufacturing using Amazon Lookout for Vision and requires as few as 30 images to bootstrap a model, builds an initial model in a couple of hours, and runs hosted inference with a returned confidence score (demo shows a test result at about 95% confidence). It detects scratches, missing components, bubbles on battery foil, and recurring process defects, supports trial detection and human verification feedback to retrain models, and integrates with S3 for dataset import.
AI-written summary · about
chapters (16)
- 0:00Introduction and cheese pizza story
- 1:11What is Amazon Lookout for Vision
- 2:17How quality inspection works today
- 5:48Problems with manual inspection
- 7:14Limitations of computer vision today
- 7:34Introducing Lookout for Vision solution
- 8:42Use cases for Lookout for Vision
- 11:25Top benefits of the service
- 13:39Customer journey and integration
- 14:46Key features and demo overview
- 15:06Live demo by Shreyas
- 20:42Early adopter success stories
- 21:26Partner ecosystem overview
- 22:10Accenture on Lookout for Vision
- 27:12Real-world food industry example
- 28:19How to get started
related talks
27mComputer vision for automated quality inspection
38mAnomaly detection in the cloud: how to detect issues in data [...]
26mServerless Architecture for Product Defect Detection
01h 00mOpening Keynote
30mMachine learning for computer vision in pharmaceutical manufacturing
15mAI as a Service: Quick and Scalable Machine Learning POCs with No-Code/Low-Code Tools
01h 06mAmazon Machine Learning to Predict Responses
36mAI-Powered Real-Time Quality Control