Recent news

I am currently looking for strong machine learning interns with expertise in few-shot learning/meta-learning, model explainability/interpretability, and robustness. Please reach out to me if you are interested in collaborating.

Feb, 2023: Work with my intern Samyadeep Basu was accepted to ICRL 2023! Our paper proposes a new algorithm to extract difficult few-shot classification tasks, with an accompanying new test benchmark.
Jan, 2023: We launched 4 exciting computer vision challenges to drive blind/low-vision assistive tech as part of the VizWiz CVPR 2023 workshop!
Aug, 2022: Some big news: I've moved to Sydney, Australia! I have transferred with Microsoft Research and will continue to collaborate with my team in the UK and new teams beyond.
Jul, 2022: We launched the VizWiz podcast with episodes featuring discussions with computer vision researchers, blind technology advocates, and industry specialists in accessibility. Podcast editing was done by @prism.media.podcasting.
Jun, 2022: We hosted the 2022 VizWiz workshop at CVPR in New Orleans! We announced the winners of our 3 challenges. The winners of the ORBIT Challenge were ORBITRON from the Australian National University and Canada-Goose from Huawei Noah's Ark Lab Canada.
May, 2022: I won the Innovating for the Future Award at the Microsoft Ability Summit 2022!
Feb, 2022: We launched the ORBIT Few-Shot Object Recognition Challenge in collaboration with the VizWiz workshop at CVPR 2022! Top 2 teams will receive $2,500 each. Closing deadline is 6 May 2022.
Dec, 2021: Participated in a panel discussion on Pursuing a Resilient and Sustainable Global Society as part of Microsoft Research's 30th Anniversary Panel Series on Inspirational and Impactful Research.
Dec, 2021: Gave a sponsor talk on the ORBIT Dataset at the Women in Machine Learning (WiML) workshop at NeurIPS 2021.
Oct, 2021: Our team won 1st prize in Microsoft's internal global hackathon out of over 6000 teams!
Oct, 2021: Gave a talk at the Microsoft Research Summit on how we can use few-shot learning to personalise AI systems to users.
Oct, 2021: We reached the 200th mentorship session milestone in the Indaba Mentorship Programme! If you're interested in mentoring African ML students, sign-up as a mentor here.
Oct, 2021: Joined the VizWiz organising committee, an annual workshop on computer vision technologies for the blind/low-vision community.
Sept, 2021: Our paper which proposes LITE was accepted to NeurIPS 2021! LITE is a general memory-efficient method for training meta-learners on large images, leading them to state-of-the-art results on the VTAB+MD and ORBIT benchmarks.
Jul, 2021: Our paper on the ORBIT dataset and benchmark was accepted to ICCV 2021! See code here to download the data and run the baseline models on the benchmark.
Jul, 2021: Our paper on collecting disability-first datasets was accepted to ASSETS 2021.
Jun, 2021: Gave a talk on the ORBIT dataset - a new real-world few-shot dataset for teachable object recognition - at the CVPR 2021 VizWiz workshop.
Apr, 2021: Wrapped up 12 months of languishing away in a global pandemic.
Mar, 2021: Was upgraded to a Senior Researcher at Microsoft Research Cambridge (and also moved to Cambridge)!
Feb, 2021: Gave a guest lecture on dataset bias at the Department of Computer Science and Technology, University of Cambridge.
Jan, 2021: Launched the Deep Learning Indaba Mentorship Programme alongside an excellent team including Aviskhar Bhoopchand, Siobhan Hall, and Kale-ab Tessera.
May, 2020: Launched the ORBIT Project, a data collection effort by blind users, with an amazing team of collaborators
Apr, 2020: Appointed as a St. Edmund's College Cambridge Postdoctoral Research Fellow
Feb, 2020: Started as a Researcher at Microsoft Research Cambridge
Oct, 2019: Defended my D.Phil thesis and passed! My examiners were Prof. Andrew Zisserman and Prof. Kristen Grauman
Sept, 2019: We wrapped up the 3rd Deep Learning Indaba in Nairobi Kenya. I coordinated the Kambule and Maathai Awards (with Adji Bousso Dieng), a How to apply for a PhD workshop (with Laura Sevilla-Lara, George Konidaris), and a computer vision parallel track (with Laura Sevilla-Lara)
Aug, 2019: Finished a 3-month machine learning internship at Microsoft Research Cambridge. I worked with Cecily Morrison, Sebastian Tschiatschek and Katja Hofmann
May, 2019: Submitted my D.Phil thesis Computer vision and natural language processing for people with vision impairment