Andrew Brown is a Member of Technical Staff at AMI Labs, where he works on advanced machine intelligence. His research spans generative models for vision, multimodal representation learning and retrieval, and the understanding of people and narratives in video. Before joining AMI, he spent more than three years as a computer-vision and machine-learning research scientist within Meta’s generative-AI organization.

Brown’s research has addressed targeted image editing, large-scale image retrieval and the analysis of people and stories across visual and audiovisual media. His publications include End-to-End Visual Editing with a Generatively Pre-Trained Artist, presented at ECCV 2022, and Smooth-AP, an ECCV 2020 method for improving large-scale image-retrieval training. He also co-authored work on movie-story retrieval, speaker recognition and automated face labelling in video archives, with the latter receiving the Best Student Paper award at MIPR 2021.

Brown earned both his DPhil in Engineering Science and his Master of Engineering from the University of Oxford. He completed his doctorate in computer vision and machine learning within Oxford’s Visual Geometry Group under Professor Andrew Zisserman. Earlier in his career, he developed machine-learning software for Bühler’s optical food-sorting systems, with his work subsequently incorporated into a new generation of the company’s machines.