Research

I work on computer vision research from an applied perspective. My work spans initial research into new methods through deployment to production. For details — including applied projects that did not result in publications, see my CV.

My research focuses on image and video retrieval problems, such copy detection and product recognition.

Much of my work concerns retrieval systems that operate at extremely large scale, and the design constraints that arise in this context.

Updates

• Jun 2024 Our report on the Video Similarity Dataset and Challenge published in CVIU [preprint]
• Jun 2023 Visual Copy Detection Workshop (VCDW) held at CVPR 2023
• Dec 2022 Visual Copy Detection Workshop (VCDW) accepted to CVPR 2023
• Dec 2022 Video Similarity Challenge launched
• Jun 2022 SSCD presented at CVPR 2022
• Dec 2021 Image Similarity Challenge presented at NeurIPS 2021

Focus areas

Copy detection

Copy detection is the task of matching modified copies of images or videos. This is a large-scale retrieval problem, where representations must be invariant to common image transformations.

Product recognition

Product recognition is the task of identifying products in images across different viewpoints and contexts.