Department of Computational and Data Sciences
Department Seminar
Speaker : Ms. Simran Khanuja
Title : “Multimodality for Multilingual NLP: The need, an application, and open questions ”
Date & Time : January 05, 2024, 03:00 PM
Venue : # 102, CDS Seminar Hall
ABSTRACT
Our world is multimodal and multilingual, demanding NLP technology to be reflective of the same. From a cognitive perspective, multimodality is intricately linked with multilingualism in the brain. From an application perspective, this would make models capable of interpreting and generating human-like responses. In this talk, we will first present a bird’s eye view on the benefits of multimodal modeling for multilingual NLP (the need). We will specifically focus on the visual and text modality, and discuss how real-world tasks would benefit from such models. Next, we will narrow down on our focus to one such application, that of translating images across cultures (an application). Specifically, we will discuss our recent work on assessing the state of generative AI to do the task, and highlight existing gaps and biases of these models. Finally, we will broaden the discussion to explore open research questions in this exciting field, and consider potential strategies for addressing these issues (open questions).
BIOGRAPHY
Simran Khanuja is a second-year PhD student at Carnegie Mellon University, working with Prof. Graham Neubig. Her research interests lie in multilingual, multimodal modeling, and downstream applications that would benefit from such models. Prior to this, she worked at Google and Microsoft Research in India for three years, contributing to the development of models and benchmarks for Indian languages.
Host Faculty: Dr. Danish Pruthi
ALL ARE WELCOME
