Neural Network Quantization Explained: FP16, INT8, PTQ and QAT
A visual guide to neural network quantization, from floating-point formats and affine quantization to PTQ, QAT and a MobileNetV2 benchmark.
ML
High-Speed Unsupervised Anomaly Detection Demo
Unsupervised anomaly detection demo for inVISION Days.
Virtual Autonomous Driving Robot Competition
Trained an agent in Gazebo to autonomously navigate a 3D obstacle course (imitation learning) and correctly read random license plates (CNN) while avoiding collisions.