Synopsis:
Multimodal information processing involves utilizing data from diverse sources such as images, videos, and text to improve real-world applications. This presentation will explore how extracting insights from multiple modalities can enhance tasks like summarization, hate speech detection, complaint mining, and medical question summarization. Combining data from videos, images, and texts can create more comprehensive summaries. The speaker will discuss their recent works in multimodal summarization, focusing on areas like comment-aware multimodal summarization, multilingual approaches, and medical question summarization. The talk will also cover the datasets and methods developed to address these challenges in detail.
Speaker(s): Dr. Vishnu S. Pendyala, Dr. Sriparna Saha
Virtual: https://events.vtools.ieee.org/m/461803