15-video ML course covers: - Fundamentals – AI/ML basics, k-NN, linear models - Classic ML – Trees, ensembles, feature engineering - Neural Nets – CNNs, RNNs, Transformers - Advanced – Vision (YOLO), Generative AI (GANs, Diffusion), Explainability Master cutting-edge techniques through practical PyTorch implementations. Perfect for both beginners and experienced practitioners looking to level up their skills.
This 12-lecture Computer Vision Technologies seminar at HSE's Faculty of Computer Science covers fundamental to advanced topics. The curriculum includes: classic approaches (OpenCV, k-NN), deep learning (CNNs, Transformers, optimization), and practical challenges (data imbalance, explainability). Students explore metric learning (Siamese nets, CLIP), generative models (GANs, diffusion), and core tasks like segmentation (UNet, SAM) and detection (YOLO, DETR). The course extends to video analysis (tracking, 3D CNNs) and production deployment (ONNX, TensorRT). Hands-on sessions use PyTorch and modern tools, providing comprehensive training in cutting-edge computer vision techniques.
This technical seminar demonstrates the end-to-end process for operationalizing ML services in production environments. Using the FindMyBike project as a practical example, we will examine: - Model conversion techniques for production deployment - API implementation strategies for model serving - gRPC server architecture for efficient inference - Cross-platform client development approaches - Operational considerations for production ML systems Participants will gain insights into transitioning machine learning models from research to reliable production services.