7 papers
Spherical Interpolation for Backward-Compatible Multimodal Representations
Simone Ricci, Niccolò Biondi, Federico Pernici
Contrastive vision-language models map visual and textual representations into a shared normalized embedding space, making cosine similarity the natural metric for cross-modal retr…
A Stationary (and Therefore Compatible) Representation is All You Need
Niccolò Biondi, Federico Pernici, Simone Ricci +1
Learning compatible representations aims to learn feature representations that can be used interchangeably over time whenever a model undergoes updates. In this paper, we demonstra…
Mitigating Negative Flips via Margin Preserving Training
Simone Ricci, Niccolò Biondi, Federico Pernici +1
Minimizing inconsistencies across successive versions of an AI system is as crucial as reducing the overall error. In image classification, such inconsistencies manifest as negativ…
-Orthogonality Regularization for Compatible Representation Learning
Simone Ricci, Niccolò Biondi, Federico Pernici +2
Retrieval systems rely on representations learned by increasingly powerful models. However, due to the high training cost and inconsistencies in learned representations, there is s…
Backward-Compatible Aligned Representations via an Orthogonal Transformation Layer
Simone Ricci, Niccolò Biondi, Federico Pernici +1
Visual retrieval systems face significant challenges when updating models with improved representations due to misalignment between the old and new representations. The costly and…
Stationary Representations: Optimally Approximating Compatibility and Implications for Improved Model Replacements
Niccolò Biondi, Federico Pernici, Simone Ricci +1
Learning compatible representations enables the interchangeable use of semantic features as models are updated over time. This is particularly relevant in search and retrieval syst…
Learning advisor networks for noisy image classification
Simone Ricci, Tiberio Uricchio, Alberto Del Bimbo
In this paper, we introduced the novel concept of advisor network to address the problem of noisy labels in image classification. Deep neural networks (DNN) are prone to performanc…