Projects
Explore the research that ELIAI team members are conducting.
Causal Knowledge Graphs for Counterfactual Claims Reasoning
![Project image Ross/Belle](/sites/default/files/styles/panel_breakpoints_theme_uoe_mobile_1x/public/thumbnails/image/ross_proj.png?itok=K39s469Y)
AI approaches to misinformation detection.
Multimodal Interpretability from Partial Sight
![Sid N project image](/sites/default/files/styles/panel_breakpoints_theme_uoe_mobile_1x/public/thumbnails/image/sid-n-project-1244x680px-gr.png?itok=wDQnbSFl)
We seek to build DGMs that capture the joint distribution over co-observed visual and language data.
Asking Your Self-Driving Car to Explain its Decisions
![Albrecht project image](/sites/default/files/styles/panel_breakpoints_theme_uoe_mobile_1x/public/thumbnails/image/albrecht-project_-900x400px-gr.png?itok=TN_YxeaZ)
Trusting an autonomous vehicle to make the right decisions.
Grounding Actions and Action Modifiers in Instructional Videos
![Keller project image](/sites/default/files/styles/panel_breakpoints_theme_uoe_mobile_1x/public/thumbnails/image/keller-project-1244x680px-gr.png?itok=jcm4l2AJ)
Grounding is the process of associating a word or a phrase with the perceptual entity it refers to
Gradient-based Learning of Complex Latent Structures
![Minervini project pic](/sites/default/files/styles/panel_breakpoints_theme_uoe_mobile_1x/public/thumbnails/image/minervini-d9d9d9-1244x690.png?itok=-cIX5GrD)
Imposing structural constraints on the latent representations learned by deep neural models
Verified Neural Network Training for Safe Human-Robot Interaction
![ELIAI Fleuriot Project Pic](/sites/default/files/styles/panel_breakpoints_theme_uoe_mobile_1x/public/thumbnails/image/fleuriot-d9d9d9-1244x680_0.png?itok=RqvV3Cer)
Robotic movement in the home and other human-centred environments requires safe, constrained behaviour.
Multimodal Integration for Sample-Efficient Deep Reinforcement Learning
![Project pic Amos Storkey PI](/sites/default/files/styles/panel_breakpoints_theme_uoe_mobile_1x/public/thumbnails/image/storkey-d9d9d9-1244x680.png?itok=fBaeXcha)
Leveraging multimodal learning to improve the transferability and sample efficiency of Reinforcement Learning (RL)