Quantifying Uncertainties of Deep Neural Networks and Its Applications

Explore a comprehensive lecture from Samsung researcher Dr. Jae Oh Woo that delves into the critical aspects of uncertainty quantification in deep neural networks and their practical applications. Learn about the methodologies and techniques used to measure and understand the uncertainties inherent in deep learning models, gaining valuable insights into how these uncertainties impact real-world applications. Discover the latest research findings and industry perspectives on managing and interpreting neural network uncertainties, with examples drawn from Samsung's practical implementations.