Current
Edge Caching for Vehicular Networks (VANET)
Learning-based content caching strategies for vehicular ad-hoc networks, integrating SDN/NDN paradigms, clustering, and spiking neural networks.
Projects
Current and past research projects. Filter by status or research theme.
Theme: 3D ReconstructionComputer VisionIntelligent Transport SystemsMachine LearningMonocular Depth EstimationPhotometric StereoReinforcement Learning
Completed
Vision, Language and Visual Retrieval
Multimodal methods connecting images and language: large-scale visual retrieval, semantic art understanding, and image caption generation.
Current
Reinforcement Learning for Traffic Signal Control
Deep and multi-agent reinforcement learning for intelligent traffic signal control, trained in rich 3D simulation and transferred to real-world networks.
Completed
Self-Supervised Monocular Depth Estimation
Learning to predict depth from single images without ground-truth supervision, with a focus on dynamic scenes and challenging conditions.
Completed
Photometric Stereo and Shape from Shading
Recovering surface normals, shape, and reflectance from images captured under varying illumination — including coloured-light, multi-spectral, and shadow-robust methods.
Completed
3D Reconstruction and Multi-View Stereo
Recovering accurate 3D shape from collections of images, using multi-view stereo, volumetric graph-cuts, and probabilistic depth-map fusion.