Asif Khan

5.19 Informatics Forum, University of Edinburgh.

I am a final-year PhD student in the Bayesian and Neural Systems research group, advised by Prof. Amos Storkey. I do research in machine learning and artificial intelligence. My PhD work focuses on deep structured latent space models with applications to disentanglement and robustness. The topics that particularly interest me include the following:

  • Deep generative models
  • Self-supervised learning
  • Bayesian inference
  • Equivariant networks and geometric deep learning

Previously, I completed MSc. in Computer Science at the University of Bonn, where I was a research assistant in the Smart Data Analytics (SDA) group. I worked on Machine Learning for Knowledge Graphs with Prof. Asja Fischer and Prof. Jens Lehmann. In my master thesis, I developed a GAN framework for speech-to-speech synthesis advised by Prof. Asja Fischer and Dr. Fabien Cardinaux. I like to work on problems with practical impact. In several of my projects, I developed ML algorithms for challenging problems such as protein design. For details, please refer to the publications.

Selected Publications

  1. Adversarial robustness of VAEs through the lens of local geometry
    Khan, Asif, and Storkey, Amos
    International Conference on Artificial Intelligence and Statistics (AISTATS) (To appear) 2023
  2. Toward real-world automated antibody design with combinatorial Bayesian optimization
    Khan, Asif, Cowen-Rivers, Alexander I, Grosnit, Antoine, Robert, Philippe A, Greiff, Victor, Smorodina, Eva, Rawat, Puneet, Akbar, Rahmad, Dreczkowski, Kamil, Tutunov, Rasul, Bou-Ammar, Dany, Wang, Jun, Storkey, Amos, and Bou-Ammar, Haitham
    Cell Reports Methods 2023
  3. HALO: HAmiltonian Latent Operators for content and motion disentanglement in image sequences
    Khan, Asif, and Storkey, Amos
    Advances in Neural Information Processing Systems (NeurIPS) 2022