322,869 research outputs found
Information-Based Learning of Deep Architectures for Feature Extraction
No full textFeature extraction is a crucial phase in complex computer vision systems. Mainly two different approaches have been proposed so far. A quite common solution is the design of appropriate filters and features based on image processing techniques, such as the SIFT descriptors. On the other hand, machine learning techniques can be applied, relying on their capabilities to automatically develop optimal processing schemes from a significant set of training examples. Recently, deep neural networks and convolutional neural networks have been shown to yield promising results in many computer vision tasks, such as object detection and recognition. This paper introduces a new computer vision deep architecture model for the hierarchical extraction of pixel-based features, that naturally embed scale and rotation invariances. Hence, the proposed feature extraction process combines the two mentioned approaches, by merging design criteria derived from image processing tools with a learning algorithm able to extract structured feature representations from data. In particular, the learning algorithm is based on information-theoretic principles and it is able to develop invariant features from unsupervised examples. Preliminary experimental results on image classification support this new challenging research direction, when compared with other deep architectures models
Semi-supervised multiclass Kernel machines with probabilistic constraints
No full textThe extension of kernel-based binary classifiers to multiclass problems has been approached with different strategies in the last decades. Nevertheless, the most frequently used schemes simply rely on different criteria to combine the decisions of a set of independently trained binary classifiers. In this paper we propose an approach that aims at establishing a connection in the training stage of the classifiers using an innovative criterion. Motivated by the increasing interest in the semi-supervised learning framework, we describe a soft-constraining scheme that allows us to include probabilistic constraints on the outputs of the classifiers, using the unlabeled training data. Embedding this knowledge in the learning process can improve the generalization capabilities of the multiclass classifier, and it leads to a more accurate approximation of a probabilistic output without an explicit post-processing. We investigate our intuition on a face identification problem with 295 classes
Learning visual features under motion invariance
No full textHumans are continuously exposed to a stream of visual data with a natural temporal structure. However, most successful computer vision algorithms work at image level, completely discarding the precious information carried by motion. In this paper, we claim that processing visual streams naturally leads to formulate the motion invariance principle, which enables the construction of a new theory of learning that originates from variational principles, just like in physics. Such principled approach is well suited for a discussion on a number of interesting questions that arise in vision, and it offers a well-posed computational scheme for the discovery of convolutional filters over the retina. Differently from traditional convolutional networks, which need massive supervision, the proposed theory offers a truly new scenario for the unsupervised processing of video signals, where features are extracted in a multi-layer architecture with motion invariance. While the theory enables the implementation of novel computer vision systems, it also sheds light on the role of information-based principles to drive possible biological solutions
Coherence constraints in facial expression recognition
No full textThis paper investigates the role of coherence constraints in recognizing facial expressions from images and video sequences. A set of constraints are introduced to bridge a pool of Convolutional Neural Networks (CNNs) during their training stage. Constraints are inspired by practical considerations on the regularity of the temporal evolution of the predictions, and by the idea of connecting the information extracted from multiple representations. We study CNNs with the aim of building a versatile recognizer of expressions in static images that can be further applied to video sequences. First, the importance of different face parts in the recognition task is studied, considering appearance and shape-related features. Then we focus on the Semi-Supervised learning setting, exploiting video data, where only a few frames are supervised. The unsupervised portion of the training data is used to enforce three types of coherence, namely temporal coherence, coherence among the predictions on the face parts and coherence between appearance and shape-based representation. Our experimental analysis shows that coherence constraints improve the quality of the expression recognizer, thus offering a suitable basis to profitably exploit unsupervised video sequences, also in cases in which some portions of the input face are not visible
Learning with convex constraints
No full textIn this paper, we focus on multitask learning and discuss the notion of learning from constraints, in which they limit the space of admissible real values of the task functions. We formulate learning as a variational problem and analyze convex constraints, with special attention to the case of linear bilateral and unilateral constraints. Interestingly, we show that the solution is not always an analytic function and that it cannot be expressed by the classic kernel expansion on the training examples. We provide exact and approximate solutions and report experimental evidence of the improvement with respect to classic kernel machines
Machine Learning: A Constraint-Based Approach
No full textMachine Learning: A Constraint-Based Approach, Second Edition provides readers with a refreshing look at the basic models and algorithms of machine learning, with an emphasis on current topics of interest that include neural networks and kernel machines. The book presents the information in a truly unified manner that is based on the notion of learning from environmental constraints. It draws a path towards deep integration with machine learning that relies on the idea of adopting multivalued logic formalisms, such as in fuzzy systems. Special attention is given to deep learning, which nicely fits the constrained-based approach followed in this book. The book presents a simpler unified notion of regularization, which is strictly connected with the parsimony principle, including many solved exercises that are classified according to the Donald Knuth ranking of difficulty, which essentially consists of a mix of warm-up exercises that lead to deeper research problems. A software simulator is also included
Neural poetry: learning to generate poems using syllables
No full textMotivated by the recent progresses on machine learning-based models that learn artistic styles, in this paper we focus on the problem of poem generation. This is a challenging task in which the machine has to capture the linguistic features that strongly characterize a certain poet, as well as the semantics of the poet’s production, that are influenced by his personal experiences and by his literary background. Since poetry is constructed using syllables, that regulate the form and structure of poems, we propose a syllable-based neural language model, and we describe a poem generation mechanism that is designed around the poet style, automatically selecting the most representative generations. The poetic work of a target author is usually not enough to successfully train modern deep neural networks, so we propose a multi-stage procedure that exploits non-poetic works of the same author, and also other publicly available huge corpora to learn syntax and grammar of the target language. We focus on the Italian poet Dante Alighieri, widely famous for his Divine Comedy. A quantitative and qualitative experimental analysis of the generated tercets is reported, where we included expert judges with strong background in humanistic studies. The generated tercets are frequently considered to be real by a generic population of judges, with relative difference of 56.25% with respect to the ones really authored by Dante, and expert judges perceived Dante’s style and rhymes in the generated text
Toward Improving the Evaluation of Visual Attention Models: A Crowdsourcing Approach
No full textHuman visual attention is a complex phenomenon. A computational modeling of this phenomenon must take into account where people look in order to evaluate which are the salient locations (spatial distribution of the fixations), when they look in those locations to understand the temporal development of the exploration (temporal order of the fixations), and how they move from one location to another with respect to the dynamics of the scene and the mechanics of the eyes (dynamics). State-of-the-art models focus on learning saliency maps from human data, a process that only takes into account the spatial component of the phenomenon and ignore its temporal and dynamical counterparts. In this work we focus on the evaluation methodology of models of human visual attention. We underline the limits of the current metrics for saliency prediction and scanpath similarity, and we introduce a statistical measure for the evaluation of the dynamics of the simulated eye movements. While deep learning models achieve astonishing performance in saliency prediction, our analysis shows their limitations in capturing the dynamics of the process. We find that unsupervised gravitational models, despite of their simplicity, outperform all competitors. Finally, exploiting a crowd-sourcing platform, we present a study aimed at evaluating how strongly the scanpaths generated with the unsupervised gravitational models appear plausible to naive and expert human observers
Friendly Training: Neural Networks Can Adapt Data To Make Learning Easier
No full textIn the last decade, motivated by the success of Deep Learning, the scientific community proposed several approaches to make the learning procedure of Neural Networks more effective. When focussing on the way in which the training data are provided to the learning machine, we can distinguish between the classic random selection of stochastic gradient-based optimization and more involved techniques that devise curricula to organize data, and progressively increase the complexity of the training set. In this paper, we propose a novel training procedure named Friendly Training that, differently from the aforementioned approaches, involves altering the training examples in order to help the model to better fulfil its learning criterion. The model is allowed to 'simplify' those examples that are too hard to be classified at a certain stage of the training procedure. The data transformation is controlled by a developmental plan that progressively reduces its impact during training, until it completely vanishes. In a sense, this is the opposite of what is commonly done in order to increase robustness against adversarial examples, i.e., Adversarial Training. Experiments on multiple datasets are provided, showing that Friendly Training yields improvements with respect to informed data sub-selection routines and random selection, especially in deep convolutional architectures. Results suggest that adapting the input data is a feasible way to stabilize learning and improve the generalization skills of the network
A language modeling-like approach to sketching
No full textSketching is a universal communication tool that, despite its simplicity, is able to efficiently express a large variety of concepts and, in some limited contexts, it can be even more immediate and effective than natural language. In this paper we explore the feasibility of using neural networks to approach sketching in the same way they are commonly used in Language Modeling. We propose a novel approach to what we refer to as “Sketch Modeling”, in which a neural network is exploited to learn a probabilistic model that estimates the probability of sketches. We focus on simple sketches and, in particular, on the case in which sketches are represented as sequences of segments. Segments and sequences can be either given – when the sketches are originally drawn in this format – or automatically generated from the input drawing by means of a procedure that we designed to create short sequences, loosely inspired by the human behavior. A Recurrent Neural Network is used to learn the sketch model and, afterward, the network is seeded with an incomplete sketch that it is asked to complete, generating one segment at each time step. We propose a set of measures to evaluate the outcome of a Beam Search-based generation procedure, showing how they can be used to identify the most promising generations. Our experimental analysis assesses the feasibility of this way of modeling sketches, also in the case in which several different categories of sketches are considered
- …
