Abstract Vision transformers (ViTs) have achieved remarkable successes across a broad range of computer vision applications. As a consequence, there has been increasing interest in extending continual learning theory and techniques to ViT architectures. We propose a new method for exemplar-free class incremental training of ViTs. The main challenge of exemplar-free continual learning is maintaining plasticity of the learner without causing catastrophic forgetting of previously learned tasks. This is often achieved via exemplar replay which can help recalibrate previous task classifiers to the feature drift which occurs when learning new tasks. Exemplar replay, however, comes at the cost of retaining samples from previous tasks which for many applications may not be possible. To address the problem of continual ViT training, we first propose gated class-attention to minimize the drift in the final ViT transformer block. This mask-based gating is applied to class-attention mechanism of the last transformer block and strongly regulates the weights crucial for previous tasks. Importantly, gated class-attention does not require the task-ID during inference, which distinguishes it from other parameter isolation methods. Secondly, we propose a new method of feature drift compensation that accommodates feature drift in the backbone when learning new tasks. The combination of gated class-attention and cascaded feature drift compensation allows for plasticity towards new tasks while limiting forgetting of previous ones. Extensive experiments performed on CIFAR-100, Tiny-ImageNet and ImageNet100 demonstrate that our exemplar-free method obtains competitive results when compared to rehearsal based ViT methods.(Code: https://github.com/OcraM17/GCAB-CFDC )

Abstract Nowadays, image-to-image translation methods, are the state of the art for the enhancement of natural images. Even if they usually show high performance in terms of accuracy, they often suffer from several limitations such as the generation of artifacts and the scalability to high resolutions. Moreover, their main drawback is the completely black-box approach that does not allow to provide the final user with any insight about the enhancement processes applied. In this paper we present a path planning algorithm which provides a step-by-step explanation of the output produced by state of the art enhancement methods. This algorithm, called eXIE, uses a variant of A $$^*$$ ∗ to emulate the enhancement process of another method through the application of an equivalent sequence of enhancing operators. We applied eXIE to explain the output of several state-of-the-art models trained on the Five-K dataset, obtaining sequences of enhancing operators able to produce very similar results in terms of performance and overcoming the huge limitation of poor interpretability of the best performing algorithms.

Photography, like painting, allows artists to express themselves through their unique style. In digital photography, this is achieved not only with the choice of the subject and the composition but also by means of post-processing operations. The automatic identification of a photographer from the style of a photo is a challenging task, for many reasons, including the lack of suitable datasets including photos taken by a diverse panel of photographers with a clear photographic style. In this paper we present PhotoStyle60, a new dataset including 5708 photographs from 60 professional and semi-professional photographers. Additionally, we selected a reduced version of the dataset, called PhotoStyle10 containing images from 10 clearly distinguishable experts. We designed the dataset to address two tasks in particular: photo authorship attribution and photographic style transfer. In the former, we conducted an extensive analysis of the dataset through several classification experiments. In the latter, we explored the potential of our dataset to transfer a photographer's style to images from the Five-K dataset. Additionally, we propose also a simple but effective multi-image style transfer method that uses multiple samples of the target style. A user study demonstrated that such a method was able to reach accurate results, preserving the semantic content of the source photograph with very few artifacts.

Social Networks represent one of the most important online sources to share content across a world-scale audience. In this context, predicting whether a post will have any impact in terms of engagement is of crucial importance to drive the profitable exploitation of these media. In the literature, several studies address this issue by leveraging direct features of the posts, typically related to the textual content and the user publishing it. In this paper, we argue that the rise of engagement is also related to another key component, which is the semantic connection among posts published by users in social media. Hence, we propose TweetGage, a Graph Neural Network solution to predict the user engagement based on a novel graph-based model that represents the relationships among posts. To validate our proposal, we focus on the Twitter platform and perform a thorough experimental campaign providing evidence of its quality.

In this paper we propose a method for Offline Handwritten Mathematical Expression recognition. The method is a fast and accurate thanks to its architecture, which include both a Convolutional Neural Network and a Recurrent Neural Network. The CNN extracts features from the image to recognize and its output is provided to the RNN which produces the mathematical expression encoded in the LATEX language. To process both sequential and non-sequential mathematical expressions we also included a deconvolutional module which, in a recursive way, segments the image for additional analysis trough a recursive process. The results obtained show a very high accuracy obtained on a large handwritten data set of 9100 samples of handwritten expressions.

The Copernicus Sentinel-1 Earth Radar Observatory, a mission funded by the European Union and developed by ESA, is a constellation of two C-band radar satellites. The satellites have been conceived to be a continuous and reliable source of C-band SAR imagery for operational applications such as mapping of global landmasses, coastal zones and monitoring of shipping routes. The Sentinel-1 satellites are built by an industrial consortium led by Thales Alenia Space Italia as Prime Contractor and with AIRBUS Defence and Space as SAR Instrument Contractor. The paper describes the general satellite architecture, AIT flow and the satellite key performances. It provides also an overview on the CSAR Instrument, its development status and prelaunch SAR performance prediction.