work READ Avatars Emotion-controllable audio-driven talking head generation using neural rendering and 3DMM. BMVC 2023 Oral. FACTS Style transfer for 3D facial animation — retargeting emotion and idiosyncratic speaking style while preserving lip sync. Eurographics 2024. TalkLoRA Personalised speech-driven facial animation using LoRA adaptors. Efficient transformer inference via chunking. BMVC 2024. DEAD High-quality video dubbing from seconds of training data using a multi-person neural rendering prior. BMVC 2025. GASP Real-time animatable Gaussian Splatting avatars from a single image or monocular video, using synthetic priors. CVPR 2025. fun
