Instructor
Katie M. Lewis

Katie M. Lewis

MIT Research Contributor

Katie M. Lewis received her PhD from MIT in Developing Domain-Specific Generative Models. As a research assistant, she developed the GIST method to generate fine-grained image-specific text descriptions using LLMs. She also developed a learning-based method to align sparse, clinical MRI brain scans with higher accuracy on 92% of subjects and 100x faster on the CPU. Some of her generative machine learning publications include At the Intersection of Conceptual Art and Deep Learning: The End of Signature, Generating Image-Specific Text for Fine-grained Object Classification, and Machine Learning for Healthcare (ML4H) at NeurIPS 2018. She recently defended her dissertation under the supervision of John Guttag and Fr├ędo Durand and has interned twice with Ira Kemelmacher-Shlizerman's team at Google.