xAILab Bamberg presents two papers at the 35th British Machine Vision Conference (BMVC) in Glasgow
Main Conference Paper: "Privacy-preserving datasets by capturing feature distributions with Conditional VAEs"
Large and well-annotated datasets are essential for advancing deep learning applications, however often costly or impossible to obtain by a single entity. In many areas, including the medical domain, approaches relying on data sharing have become critical to address those challenges. While effective in increasing dataset size and diversity, data sharing raises significant privacy concerns. Commonly employed anonymization methods based on the k-anonymity paradigm often fail to preserve data diversity, affecting model robustness. This work introduces a novel approach using Conditional Variational Autoencoders (CVAEs) trained on feature vectors extracted from large pre-trained vision foundation models. Foundation models effectively detect and represent complex patterns across diverse domains, allowing the CVAE to faithfully capture the embedding space of a given data distribution to generate (sample) a diverse, privacy-respecting, and potentially unbounded set of synthetic feature vectors. Our method notably outperforms traditional approaches in both medical and natural image domains, exhibiting greater dataset diversity and higher robustness against perturbations while preserving sample privacy. These results underscore the potential of generative models to significantly impact deep learning applications in data-scarce and privacy-sensitive environments.
For more details, you can access the paper here and the associated code repository here.
RROW-Workshop Paper: "Unsupervised Feature Orthogonalization for Learning Distortion-Invariant Representations"
We introduce unORANIC+, a novel method that integrates unsupervised feature orthogonalization with the ability of a Vision Transformer to capture both local and global relationships for improved robustness and generalizability. The streamlined architecture of unORANIC+ effectively separates anatomical and image-specific attributes, resulting in robust and unbiased latent representations that allow the model to demonstrate excellent performance across various medical image analysis tasks and diverse datasets. Extensive experimentation demonstrates unORANIC+'s reconstruction proficiency, corruption resilience, as well as capability to revise existing image distortions. Additionally, the model exhibits notable aptitude in downstream tasks such as disease classification and corruption detection. We confirm its adaptability to diverse datasets of varying image sources and sample sizes which positions the method as a promising algorithm for advanced medical image analysis, particularly in resource-constrained environments lacking large, tailored datasets.
For more details, you can access the paper here and the associated code repository here.
About BMVC and RROW
The British Machine Vision Conference (BMVC) is the British Machine Vision Association's (BMVA) annual conference on machine vision, image processing and pattern recognition. It is one of the most important international conferences on computer vision and related fields held in the UK. With increasing popularity and quality, it has become a respected event in the machine vision calendar.
The RROW workshop addresses the challenges faced by deep neural networks in unexpected scenes and environments that differ from the training data. The focus will be on methods and datasets to improve the robustness and adaptability of AI for critical applications such as automated driving, robotics and medical imaging.
This year, the conference will take place in Glasgow at the Scottish Event Campus, a renowned venue known for its modern facilities and picturesque location on the River Clyde.
