Generative Models: Trajectory Forecasting

This lecture covers generative models for trajectory forecasting in autonomous vehicles, focusing on discriminative vs generative models, the importance of generative models, and the use of VAES and GANS. It delves into PixelCNN, VAES, GANS, ProGAN, StyleGAN, CGANS, and case studies. The instructor discusses the training and evaluation process, evaluation metrics, and various case studies related to pedestrian image generation and shared representations for driving simulators.