expectopatronm / Rommelag-Defect-Identification-Challenge

Defect Identification Challenge to find an efficient and exact process for error detection in the production line using Computer Vision, Deep Learning or Classical Machine Learning based techniques.

Industrial Error Detection with Machine Learning Hackathon

PionierGarage | Entrepreneurs. KIT

2nd Place Winner

The challenge is to find an efficient and exact process for error detection in the production line. You will get real data of a manufacturing defect of an operational manufacturing with the aim to classify it in the team binary (good / bad). The current approach is a visual inspection of the products by a camera. This procedure, as well as the final processing (deep learning / feature extraction / ...) is up to you as a participant!

Proposed Architecture

Performance Metrics