Research

My research is centered around ensuring the practical reliability and efficiency of neural networks in real-world settings. I apply optimization techniques to achieve neural network quantization and pruning, aiming to optimize their performance while reducing computational demands. Moreover, I'm dedicated to enhancing the robustness and trustworthiness of large neural networks, addressing critical aspects like bias and interpretability. At Granular AI, my focus lies in the development of tools tailored for geospatial vision and language models, with the aim of leveraging AI advancements to analyze and interpret geospatial data more effectively.