Ashhar Aziz is an AI researcher specializing in multimodal learning, meta-learning, and AI-generated content detection. His work on vision- language models, adversarial robustness, and deep learning has been recognized at top research venues. He was invited to the Google DeepMind Research Symposium and serves as an Associate Organizer for the Defactify 4 Workshop at AAAI-25. Ashhar has also secured top ranks in AI challenges like SCSPAR at NCVPRIPG’24 and INTERSPEECH’24.

Previously, he researched continual learning for vision-language models at Eindhoven University of Technology and AI-generated content detection at the University of South Carolina, collaborating with mentors from Amazon and Meta US. His work extends to text-image person re- identification at IIIT-Delhi and speaker- language diarization at MIDAS Labs. With papers under review at leading AI conferences, he is focused on developing novel evaluation metrics and tools to enhance AI's trustworthiness in real-world applications.