ongoing Efficient Multimodal Reasoning AIRR-funded research on reasoning efficiently across vision, language, and other modalities. Perfect Recollection for Clearer Insight Efficient foundation models and vision-language models for long-context perception (EPSRC). previous COG-MHEAR: Multimodal Hearing Aids Cognitively-inspired, 5G-IoT multimodal hearing aids with RF-guided fusion (EPSRC). Continual Learning for Audio Sensing Multi-task and continual learning for audio sensing on resource-constrained platforms (EPSRC). OPERA: Opportunistic Passive Radar AI-driven RF sensing for opportunistic passive radar and contextual sensing (EPSRC).
