معرفی
Aishwarya Agrawal is an Assistant Professor in the Department of Computer Science and Operations Research at the University of Montreal and a core academic member of Mila – Quebec AI Institute, where she holds the Canada CIFAR AI Chair. She also works as a Research Scientist at Google DeepMind, spending one day per week there. Her research sits at the intersection of computer vision, natural language processing, and deep learning, aiming to build AI systems that can both 'see' and 'talk'.
Her research interests include vision-language reasoning, data-efficient pretraining, robust evaluation, improving model reliability and explainability, and aligning AI systems with diverse cultural values. She has made significant contributions to visual question answering, particularly in addressing dataset bias through work like the VQA-CP dataset.
The recent publications show a strong focus on evaluating and improving vision-language models, especially in areas like cultural alignment, consistency, reliability, and robustness. There is a clear trend toward developing better evaluation methodologies and enhancing model behavior in complex, real-world scenarios. Her group actively publishes in top-tier conferences including CVPR, NeurIPS, ICML, and ACL.
- Canada CIFAR AI Chair Award
- Georgia Tech 2020 Sigma Xi Best Ph.D. Thesis Award
- Georgia Tech 2020 College of Computing Dissertation Award
- Runner-up, 2019 AAAI/ACM SIGAI Dissertation Award
- NVIDIA Graduate Fellowship 2018-2019
- Rising Stars in EECS 2018
Aishwarya advises several PhD and MSc students, including Rabiul Awal, Qian Yang, Le Zhang, and Oscar Mañas. She has received research grants, such as one from Samsung for compositional reasoning in multimodal transformers. She is actively recruiting graduate students and postdocs through Mila’s supervision request process.
She co-organized the annual VQA Challenge and Workshop from 2016 to 2021 and is involved in broader AI community initiatives, including panels and talks on leadership and AI research trends. Her lab, part of Mila, fosters a collaborative environment focused on advancing multimodal AI.
