BMVA Symposium on Continual AI: Learning to Adapt in a Continuous World.

One Day Meeting: BMVA Symposium on Continual AI: Learning to Adapt in a Continuous World.

Wednesday 24 June 2026

Chairs: Paris Giampouras (University of Warwick), Haoran Ni (University of Warwick), Julio Hurtado (University of Warwick)

Please register to attend on this link:    Register Here to Attend

Queries? contact the Meeting's Organiser Andrew Gilbert here

Invited Speakers

Overview

We invite the community to attend the symposium and present their work that challenges the idea of static models by focusing on dynamic environments. Presentations may include both ongoing and published research that demonstrates model adaptation through modifications or adjustments to original weights. This can involve learning new concepts or tasks, effectively reusing previously learned weights, forgetting sensitive information or biases, and merging different sources of knowledge. The meeting aims to foster interdisciplinary dialogue and collaboration to support the responsible translation of AI technologies into real-world scenarios.

Schedule

Start   End   Title
09:45   10:00   Registration & Coffee
10:00   10:45   Invited Keynote Speaker – Jun Wang (University College London)
10:45   11:30   Accepted Talks Pt. 1
11:30   12:00   Coffee Break
12:00   13:00   Accepted Talks Pt. 2
13:00   14:00   Lunch
14:00   14:45   Invited Keynote Speaker – Georgina Cosma (Loughborough University)
14:45   15.15   Coffee Break
15:15   16:00   Invited Keynote Speaker – Amos Storkey (University of Edinburgh)
16:00   16:45   Accepted Talks Pt. 3
16:45   17:00   Closing Remarks

Accepted Talks

12 min each + 3 mins questions

Accepted Talks Part 1:

Luca Manneschi - The University of Sheffield - From Signal Drift to Meta-Learning: Continual Adaptation in Neuromuscular Human–Machine Interfaces

Xin Feng - University of Edinburgh - Restoring without Forgetting: Filter-level Continual Adaptation via Integrated Gradient Attribution

Adrian Bors - University of York - Continual learning by means of generative A

Accepted Talks Part 2:

Can Peng - University of Oxford - Multi-Label Federated Learning under Label-Distribution Skew

Mamatha Thota - University of Lincoln - Stable Drift: A Patient-Aware Replay Method for Continual Learning in Medical Imaging

Abhiroop Chatterjee - Jadavpur University, Rochester Institute of Technology - CASPA: Graph-Structured Concept Anchors for Modality-Agnostic Adaptation in Vision–Language Models

Anoushka Harit - University of Cambridge - EWC-Guided Diffusion Replay for Exemplar-Free Continual Learning in Medical Imaging

Accepted Talks Part 3:

Vishal Thengane- University of Surrey - COPE: Scene-Contextualized Incremental Few-Shot 3D Segmentation

Zhaohan Zhang - Queen Mary University of London - Get confused cautiously: Textual sequence memorization erasure with selective entropy maximization

Anjie Le - University of Oxford - POUR: A Provably Optimal Method for Unlearning Representation via Neural Collapse

Important: This is an in-person event, with no virtual attendance option. We kindly ask all presenters to join us at the British Computer Society on the day.

Meeting Location

The meeting will take place at:

British Computer Society (BCS), 25 Copthall Avenue, London EC2R 7BP

Registration

We keep the cost of attending these events as low as possible to ensure no barriers from the whole computer vision community attending. The registration costs are as follows

To attend please register via charitysuite on this link:    Register Here