Research
All publications on Google Scholar.
Publications
- 2026
- Advancing Conversational Diagnostic AI with Multimodal Reasoning Nature Medicine. Nature Medicine
- Towards Conversational Medical AI with Eyes, Ears and a Voice Preprint. arxiv
- 2025
- Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long context, and Next Generation Agentic Capabilities Technical report. arxiv
- Advancing Geometry with AI: Multi-agent Generation of Polytopes Preprint. arxiv
- 2024
- Collaboration Between Clinicians and Vision–Language Models in Radiology Report Generation Nature Medicine. Nature Medicine
- Gemini 1.5: Unlocking Multimodal Understanding Across Millions of Tokens of Context Technical report. PDF
- Advancing Multimodal Medical Capabilities of Gemini Preprint. arxiv
- Capabilities of Gemini Models in Medicine Preprint. arxiv
- 2023
- Consensus, Dissensus and Synergy Between Clinicians and Specialist Foundation models in Radiology Report Generation Preprint. arxiv
- Self-Supervised Video Pretraining Yields Robust and More Human-Aligned Visual Representations Neural Information Processing Systems (NeurIPS). arxiv
- 2022
- From Data to Functa: Your Data Point is a Function and You Can Treat It Like One International Conference on Machine Learning (ICML). arxiv
- Towards Real Robot Learning in the Wild: A Case Study in Bipedal Locomotion Conference on Robot Learning (CoRL). JMLRSupplementary material
- 2021
- Inferring a Continuous Distribution of Atom Coordinates from Cryo-EM Images Using VAEs Neural Information Processing Systems (NeurIPS), Workshop on Machine Learning in Structural Biology. arxiv
- Multimodal Few-Shot Learning with Frozen Language Models Neural Information Processing Systems (NeurIPS). arxiv
- From Motor Control to Team Play in Simulated Humanoid Football Science Robotics. Science Roboticsarxiv
- Generative Art Using Neural Visual Grammars and Dual Encoders Preprint. arxiv
- 2020
- Game Plan: What AI can do for Football, and What Football can do for AI Journal of Artificial Intelligence Research (JAIR). JAIRarxiv
- Contrastive Training for Improved Out-of-Distribution Detection Preprint. arxiv
- PolyGen: An Autoregressive Generative Model of 3D Meshes International Conference on Machine Learning (ICML). arxiv
- 2019
- Unsupervised Doodling and Painting with Improved SPIRAL Preprint. Interactive Overviewarxiv
- A Hierarchical Probabilistic U-Net for Modeling Multi-Scale Ambiguities Preprint. arxiv
- Data-Efficient Image Recognition with Contrastive Predictive Coding International Conference on Machine Learning (ICML). arxiv
- Meta-Learning surrogate models for sequential decision making Preprint. arxiv
- 2018
- Generating Diverse Programs with Instruction Conditioned Reinforced Adversarial Learning Preprint. arxiv
- Attentive Neural Processes Neural Information Processing Systems (NeurIPS), Workshop on Bayesian Deep Learning. arxiv
- Neural Processes International Conference on Machine Learning (ICML), Workshop on Theoretical Foundations and Applications of Deep Generative Models. arxiv
- Conditional Neural Processes International Conference on Machine Learning (ICML). arxivCode
- Encoding Spatial Relations from Natural Language Preprint. arxiv
- Learning models for visual 3D localization with implicit mapping Neural Information Processing Systems (NeurIPS), Workshop on Bayesian Deep Learning. arxiv
- Consistent Jumpy Predictions for Videos and Scenes Neural Information Processing Systems (NeurIPS), Workshop on Bayesian Deep Learning. arxiv
- A Probabilistic U-Net for Segmentation of Ambiguous Images Neural Information Processing Systems (NeurIPS). arxiv
- Neural Scene Representation and Rendering Science. OverviewVideoScience PDFPreprint PDFDatasets
- Generative Temporal Models with Spatial Memory for Partially Observed Environments International Conference on Machine Learning (ICML). arxiv
- Synthesizing Programs for Images using Reinforced Adversarial Learning International Conference on Machine Learning (ICML). arxivOverviewVideoSlidesPoster
- Machine Theory of Mind International Conference on Machine Learning (ICML). arXiv
- Kickstarting Deep Reinforcement Learning Preprint. arXiv
- Learning and Querying Fast Generative Models for Reinforcement Learning Preprint. arXiv
- 2017
- The Multi-Entity Variational Autoencoder Neural Information Processing Systems (NIPS), Workshop on Learning Disentangled Features. PDF
- Few-shot Autoregressive Density Estimation: Towards Learning to Learn Distributions International Conference on Learning Representations (ICLR). arXiv
- Emergence of Locomotion Behaviours in Rich Environments Preprint. arXivBlogVideos
- 2016
- Unsupervised Learning of 3D Structure from Images Neural Information Processing Systems (NIPS). arXivVideos
- Attend, Infer, Repeat: Fast Scene Understanding with Generative Models Neural Information Processing Systems (NIPS). PDFOverview pagearXivVideo MP4Video StreamCode (thanks to Adam Kosiorek)
- 2015
- Reinforced Variational Inference Neural Information Processing Systems (NIPS), Workshop on Advances in Approximate Bayesian Inference. Oral presentation. PDFPoster
- Kernel-Based Just-In-Time Learning for Passing Expectation Propagation Messages Uncertainty in Artificial Intelligence (UAI). arXivCodePoster
- Just-In-Time Kernel Regression for Expectation Propagation International Conference on Machine Learning (ICML), Workshop on Large Scale Kernel Learning. PDFCode
- Consensus Message Passing for Layered Graphical Models International Conference on Artificial Intelligence and Statistics (AISTATS). PDFSupplementary materialarXiv
- 2014
- Just-In-Time Learning for Fast and Flexible Inference Neural Information Processing Systems (NIPS). PDFSupplementary materialPoster
- The PASCAL Visual Object Classes Challenge — a Retrospective International Journal of Computer Vision (IJCV). Springer PDFPreprint PDF
- 2013
- Assessing the Significance of Performance Differences on the PASCAL VOC Challenges via Bootstrapping Technical note. PDF
- The Shape Boltzmann Machine: a Strong Model of Object Shape International Journal of Computer Vision (IJCV). CVPR special issue. Springer PDFOverview pagePreprint PDF
- 2012
- A Generative Model for Parts-based Object Segmentation Neural Information Processing Systems (NIPS). PDFCodePoster
- The Shape Boltzmann Machine: a Strong Model of Object Shape Computer Vision and Pattern Recognition (CVPR). Oral presentation. PDFOverview pageCodeVideo
- 2011
- Factored Shapes and Appearances for Parts-based Object Understanding British Machine Vision Conference (BMVC). Oral presentation. PDFAbstractSupplementary materialPoster
Theses
- Generative Probabilistic Models for Object Segmentation The University of Edinburgh, PhD thesis, 2013. PDF
- Evolving Robust Control Strategies for Simulated Animats The University of Edinburgh, BEng dissertation, 2009. PDFOverview pagePoster
Writing
- 6 Things Everyone Should Know About Visualizing Research Results January 2015
- The Zeroth Rule of Presentations October 2012
- Patterns for Research in Machine Learning July 2012
- Thoughts on the Role of Supervision in Computer Vision May 2011
