Connecting Vision and Language End-to-End - 2018

Explore cutting-edge research in connecting vision and language through this 1-hour lecture by Kate Saenko from Boston University. Delve into End-to-End Module Networks (N2NMNs), a novel approach for answering questions about images by decomposing them into modular sub-problems. Learn how these networks achieve better accuracy than state-of-the-art attentional approaches while providing interpretable architectures. Discover recent advancements in dense video captioning, including an end-to-end network that localizes and describes activities in long videos. Gain insights into the broader field of Artificial Intelligence, focusing on Adaptive Machine Learning, Vision and Language Understanding, and Deep Learning from an expert in the field.