Scholar
Puneet K. Dokania
Google Scholar ID: WsM7ybkAAAAJ
University of Oxford | Bosch (Five AI)
Citations & Impact
Citations
7,187
H-index
28
i10-index
42
Publications
20
Co-authors
115
list available
Contact
Resume (English only)
Academic Achievements
- Published multiple papers covering areas such as neural network robustness, incremental learning, neural network discretization/pruning, generative models, vision and language, and more. Served as a Programme Chair at BMVC 2022 (The 33rd British Machine Vision Conference).
Research Experience
- At the University of Oxford, he primarily supervises PhD students working on topics related to robustness, continual learning, and efficient methods for neural network training. At Five AI Oxford office, he leads a small team of research scientists, helps shape the company's research agenda related to the perception side of their underlying AI systems, and assists the applied engineering team in building research-based products.
Education
- During his PhD, he was part of the CVN group at INRIA and CentraleSupélec, supervised by Prof. M. Pawan Kumar and Prof. Nikos Paragios from October 2012 to March 2016.
Background
- Senior researcher in machine learning and computer vision. Currently working at the Torr Vision Group (University of Oxford) and serving as a research director at Five AI Oxford. Main research interests include optimization, deep learning, generative models, and their applications to vision and language.
Miscellany
- Collaborated with many brilliant people from institutions such as UC Berkeley, INRIA, Adobe Research, Meta, IIIT Hyderabad, and more.
