Recorded Talks

Opening Remarks

Understanding the robustness of Vision Transformers

Few-shot Classification by Recycling Deep Learning

Bayesian Few-Shot Classification with One-vs-Each Pólya-Gamma Augmented Gaussian Processes

Low-shot Scene Decomposition via Reconstruction

Imagination Supervised Machines that can See, Create, Drive and Feel

Naturally Limited Videos of Fine-Grained Actions - How to Collect and Learn from Long-Tailed Videos

Data Efficient Learning for Granular Scene Understanding

Contributed talks 1 & 2

contributed talks 3 & 4

Closing Remarks