Abstract
Convolutional Neural Networks (CNN) have become the gold standard in many visual recognition tasks including medical applications. Due to their high variance, however, these models are prone to over-fit the data they are trained on. To mitigate this problem, one of the most common strategies, is to perform data augmentation. Rotation, scaling and translation are common operations. In this work we propose an alternative method to rotation-based data augmentation where the rotation transformation is performed inside the CNN architecture. In each training batch the weights of all convolutional layers are rotated by the same random angle. We validate our proposed method empirically showing its usefulness under different scenarios.

Abstract
We propose a framework to model the distribution of sequential data coming from a set of entities connected in a graph with a known topology. The method is based on a mixture of shared hidden Markov models (HMMs), which are jointly trained in order to exploit the knowledge of the graph structure and in such a way that the obtained mixtures tend to be sparse. Experiments in different application domains demonstrate the effectiveness and versatility of the method.

Abstract
Rat’s gait analysis plays an important role in the assessment of the impact of certain drugs on the treatment of osteoarthritis. Since movement-evoked pain is an early characteristic of this degenerative joint disease, the affected animal modifies its behavior to protect the injured joint from load while walking, altering its gait’s parameters, which can be detected through a video analysis. Because commercially available video-based gait systems still present many limitations, researchers often choose to develop a customized system for the acquisition of the videos and analyze them manually, a laborious and time-consuming task prone to high user variability. Therefore, and bearing in mind the recent advances in machine learning and computer vision fields, as well as their presence in many tracking and recognition applications, this work is driven by the need to find a solution to automate the detection and quantification of the animal’s gait changes making it an easier, faster, simpler and more robust task. Thus, a comparison between different methodologies to detect and segment the animal under degraded luminance conditions is presented in this paper as well as an algorithm to detect, segment and classify the animal’s paws. © 2019, Springer Nature Switzerland AG.

Abstract
Traditional image segmentation algorithms operate by iteratively working over an image, as if refining a segmentation until a stopping criterion is met. Deep learning has replaced traditional approaches, achieving state-of-the-art performance in many problems, one of them being image segmentation. However, the concept of segmentation refinement is not present anymore, since usually the images are segmented in a single step. This work focuses on the refinement of image segmentations using deep convolutional neural networks, with the addition of a quality prediction output. The output from a state-of-the-art base segmenter is refined, simultaneously improving it and trying to predict its quality. We show that the quality concept can be used as a regularizer while training a network for direct segmentation refinement.

Abstract
The Overhead Power Distribution Lines present a wide range of insulator components, which have different shapes and types of building materials. These components are usually exposed to weather and operational conditions that may cause deviations in their shapes, colors or textures. These changes might hinder the development of automatic systems for visual inspection. In this perspective, this work presents a robust methodology for image classification, which aims at the efficient distribution insulator class identification, regardless of its degradation level. This work can be characterized by the following steps: implementation of Convolutional Neural Network (CNN); transfer learning; attribute vector acquisition and design of hybrid classifier architectures to improve the discrimination efficiency. In summary, a previously trained CNN goes through a fine tuning stage for later use as a feature extractor for training a new set of classifiers. A comparative study was conducted to identify which classifier architecture obtained the best discrimination performance for non-conforming components. The proposed methodology showed a significant improvement in classification performance, obtaining 95% overall accuracy in the identification of non-conforming component classes. © 2019,IEEE

Abstract
Despite the high performance of current presentation attack detection (PAD) methods, the robustness to unseen attacks is still an under addressed challenge. This work approaches the problem by enforcing the learning of the bona fide presentations while making the model less dependent on the presentation attack instrument species (PAIS). The proposed model comprises an encoder, mapping from input features to latent representations, and two classifiers operating on these underlying representations: (i) the task-classifier, for predicting the class labels (as bona fide or attack); and (ii) the species-classifier, for predicting the PAIS. In the learning stage, the encoder is trained to help the task-classifier while trying to fool the species-classifier. Plus, an additional training objective enforcing the similarity of the latent distributions of different species is added leading to a 'PAIspecies'- independent model. The experimental results demonstrated that the proposed regularisation strategies equipped the neural network with increased PAD robustness. The adversarial model obtained better loss and accuracy as well as improved error rates in the detection of attack and bona fide presentations. © 2019 Gesellschaft fur Informatik (GI). All rights reserved.