1 code implementation • 15 Jan 2023 • Federico Pernici, Matteo Bruni, Claudio Baecchi, Alberto del Bimbo
Convolutional Neural Networks (CNNs) trained with the Softmax loss are widely used classification models for several vision tasks.
1 code implementation • 16 Nov 2022 • Niccolo Biondi, Federico Pernici, Matteo Bruni, Daniele Mugnai, Alberto del Bimbo
We identify stationarity as the property that the feature representation is required to hold to achieve compatibility and propose a novel training procedure that encourages local and global stationarity on the learned representation.
1 code implementation • 11 May 2022 • Tommaso Barletti, Niccolo' Biondi, Federico Pernici, Matteo Bruni, Alberto del Bimbo
In this paper, we propose a novel training procedure for the continual representation learning problem in which a neural network model is sequentially learned to alleviate catastrophic forgetting in visual search tasks.
1 code implementation • 15 Nov 2021 • Niccolo Biondi, Federico Pernici, Matteo Bruni, Alberto del Bimbo
Compatible features enable the direct comparison of old and new learned features allowing to use them interchangeably over time.
1 code implementation • IEEE Transactions on Neural Networks and Learning Systems 2021 • Federico Pernici, Matteo Bruni, Claudio Baecchi, Alberto del Bimbo
Typically, a learnable transformation (i. e. the classifier) is placed at the end of such models returning a value for each class used for classification.
1 code implementation • 16 Oct 2020 • Federico Pernici, Matteo Bruni, Claudio Baecchi, Francesco Turchini, Alberto del Bimbo
Contrarily to the standard expanding classifier, this allows: (a) the output nodes of future unseen classes to firstly see negative samples since the beginning of learning together with the positive samples that incrementally arrive; (b) to learn features that do not change their geometric configuration as novel classes are incorporated in the learning model.
no code implementations • 27 Feb 2019 • Federico Pernici, Matteo Bruni, Claudio Baecchi, Alberto del Bimbo
Typically, a learnable transformation (i. e. the classifier) is placed at the end of such models returning a value for each class used for classification.
no code implementations • CVPR 2018 • Federico Pernici, Federico Bartoli, Matteo Bruni, Alberto del Bimbo
It is shown that the proposed learning procedure is asymptotically stable and can be effectively used in relevant applications like multiple face identification and tracking from unconstrained video streams.