Internet Explorer | EPFL Graph Search

Internet Explorer, parfois abrégé IE, est un ancien navigateur web développé par la société américaine Microsoft et installé par défaut avec Windows. Entre la fin des années 1990, quand il détrône Netscape Navigator, jusque vers 2012, c'est le navigateur web le plus utilisé au monde atteignant plus de 95 % de part de marché en 2004. Internet Explorer atteint moins de 1% de part de marché en mars 2021, et n'est plus activement développé, Microsoft l'ayant remplacé par Microsoft Edge développé depuis 2015 et initialement basé sur EdgeHTML puis, depuis 2020, sur Chromium. Internet Explorer Mobile a été développé pour les Windows Phone. Microsoft annonce l'arrêt de son support en 2015 en raison d'une utilisation relativement faible depuis la version 6. La version 11 du navigateur cohabitera dans Windows 10 avec Microsoft Edge. Depuis 2019, Microsoft conseille de ne plus utiliser Internet Explorer. Chris Jackson, expert en cybersécurité de Windows chez Microsoft, encourage les utilisateu

Person Retrieval in Surveillance Videos Via Deep Attribute Mining and Reasoning

Peixia Li, Zhiye Wang, Zhen Wei

Person retrieval largely relies on the appearance features of pedestrians. This task is rather more difficult in surveillance videos due to the limitations of extracting robust appearance features brought by the cross-view and cross-camera data with lower image resolution, motion blur, occlusion and other kinds of image degradation. To build up a more reliable person retrieval system, recent works introduced appearance attribute models to describe and distinguish different persons with high-level semantic concepts. Despite the progress of previous works, the value of utilizing appearance attributes is still under-explored. On one hand, existing methods lack for concise and precise attribute representations that are specific for each attribute category and, in the meantime, are able to filter noisy information in irrelevant spatial locations and useless patterns. On the other hand, correlation and reasoning between different attributes are neglected, which could generate more useful information and add more robustness to the retrieval system. In this paper, we propose an Attribute Mining and Reasoning (AMR) framework which is capable to handle the issues in question. The AMR makes better use of appearance attributes with two main components. First, the AMR disentangles the representations of different attributes by localizing their spatial positions and identifying their effective patterns in a weakly supervised manner. To achieve more reliable localization, we propose the Attribute Localization Ensemble (ALE) module that is consisted of multiple localization heads and a voting mechanism. Second, we introduce the Attribute Reasoning (AR) module to correlate different attributes together with the global appearance features and discover their latent relations to generate more comprehensive descriptions of pedestrians. Extensive experiments on DukeMTMC-ReID and Market-1501 datasets demonstrate the effectiveness of the proposed AMR framework as well as its superiority over the existing state-of-the-art methods. The AMR model also shows great generalization ability on the unseen CUHK03 dataset when it is only trained on Market-1501 dataset.

IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC2021

On Vacuous and Non-Vacuous Generalization Bounds for Deep Neural Networks.

Konstantinos Pitas

Deep neural networks have been empirically successful in a variety of tasks, however their theoretical understanding is still poor. In particular, modern deep neural networks have many more parameters than training data. Thus, in principle they should overfit the training samples and exhibit poor generalization to the complete data distribution. Counter intuitively however, they manage to achieve both high training accuracy and high testing accuracy. One can prove generalization using a validation set, however this can be difficult when training samples are limited and at the same time we do not obtain any information about why deep neural networks generalize well. Another approach is to estimate the complexity of the deep neural network. The hypothesis is that if a network with high training accuracy has high complexity it will have memorized the data, while if it has low complexity it will have learned generalizable patterns. In the first part of this thesis we explore Spectral Complexity, a measure of complexity that depends on combinations of norms of the weight matrices of the deep neural network. For a dataset that is difficult to classify, with no underlying model and/or no recurring pattern, for example one where the labels have been chosen randomly, spectral complexity has a large value, reflecting that the network needs to memorize the labels, and will not generalize well. Putting back the real labels, the spectral complexity becomes lower reflecting that some structure is present and the network has learned patterns that might generalize to unseen data. Spectral complexity results in vacuous estimates of the generalization error (the difference between the training and testing accuracy), and we show that it can lead to counterintuitive results when comparing the generalization error of different architectures. In the second part of the thesis we explore non-vacuous estimates of the generalization error. In Chapter 2 we analyze the case of PAC-Bayes where a posterior distribution over the weights of a deep neural network is learned using stochastic variational inference, and the generalization error is the KL divergence between this posterior and a prior distribution. We find that a common approximation where the posterior is constrained to be Gaussian with diagonal covariance, known as the mean-field approximation, limits significantly any gains in bound tightness. We find that, if we choose the prior mean to be the random network initialization, the generalization error estimate tightens significantly. In Chapter 3 we explore an existing approach to learning the prior mean, in PAC-Bayes, from the training set. Specifically, we explore differential privacy, which ensures that the training samples contribute only a limited amount of information to the prior, making it distribution and not training set dependent. In this way the prior should generalize well to unseen data (as it hasn't memorized individual samples) and at the same time any posterior distribution that is close to it in terms of the KL divergence will also exhibit good generalization.

EPFL2020

The link between innovation and finance

Cristiano Zazzara

This thesis comprises three essays on the relationship between innovation and finance. Although previous research has acknowledged the multi-faceted nature of innovation, this thesis unpacks its constituent elements, compares alternative drivers of innovation — focusing on the role of financial variables —, and explores when each of these drivers is important relative to the others. Evidence is based on survey data on Italian manufacturing firms over the years 1995-2003. The main contribution of this dissertation is to show how innovation at the firm level varies across different dimensions, documenting several robust empirical regularities, which must be accounted for. The first essay explores the determinants of various characteristics on the propensity to innovate in the Italian manufacturing sector. First, we concentrate on firms characteristics, and include in the study age of firm, size of firm, export shares of revenues and group membership. Second, we examine industry characteristics, such as type of industry in terms of technology level and market concentration. We also test the impact of characteristics of the firm's geographical area, including an indicator of financial development for the firm region. We finally extend our model including financial variables at the firm level in order to evaluate the association between profitability & risk and innovation. The second essay further examines the determinants of innovation, focusing on why firms develop product innovations rather than process innovation, or product and process innovations simultaneously. Empirical studies have highlighted the different impact of product and process innovation on various factors (such as, for example, international competitiveness, the level of employment and the types of skills used, and on the profit rate of firms). However, little is known from either a theoretical or an empirical point of view about the determinants of these different types of innovation. The third essay examines the question of whether access to credit is empirically more difficult for innovative firms than for those undertaking traditional investment projects. The idea that market financing of innovative projects is likely to encounter severe obstacles is not new. Indeed, it dates back at least to Schumpeter's defence of monopoly power since it not only guarantees that innovative firms can internalise the benefits of their innovations, but also provides funding for future innovative projects. Testing this proposition requires information on the innovative content of a given firm and on the limits to borrowing it encounters. Using the above mentioned survey data, we are able to identify credit constrained firms and use this information to make inferences on bank's lending strategies towards firms whose investment projects differ according to expected return and riskiness.

EPFL2008