In this paper, regularized lightweight deep convolutional neural network models, capable of effectively operating in realtime on devices with restricted computational power for highresolution video input are proposed....
详细信息
ISBN:
(纸本)9781728166629
In this paper, regularized lightweight deep convolutional neural network models, capable of effectively operating in realtime on devices with restricted computational power for highresolution video input are proposed. Furthermore, a novel regularization method motivated by the Quadratic Mutual Information, in order to improve the generalization ability of the utilized models is proposed. Extensive experiments on various binary classification problems involved in autonomous systems are performed, indicating the effectiveness of the proposed models as well as of the proposed regularizer.
We propose an approach to retrospective change-point estimation that includes learning feature representations from data. The feature representations are specified within a differentiable programming framework, that i...
详细信息
ISBN:
(纸本)9781728166629
We propose an approach to retrospective change-point estimation that includes learning feature representations from data. The feature representations are specified within a differentiable programming framework, that is, as parameterized mappings amenable to automatic differentiation. The proposed method uses these feature representations in a penalized least-squares objective into which known change-point labels can be incorporated. We propose to minimize the objective using an alternating optimization procedure. We present numerical illustrations on synthetic and real data showing that learning feature representations can result in more accurate estimation of change-point locations.
We propose a new framework for Detection/Estimation designed to avoid the loss of salient information in the process of reducing the dimensionality of digitized data. The main idea is a Semi-Supervised learning pre-pr...
详细信息
ISBN:
(纸本)9781424423750
We propose a new framework for Detection/Estimation designed to avoid the loss of salient information in the process of reducing the dimensionality of digitized data. The main idea is a Semi-Supervised learning pre-processing scheme ([1]) based oil Compressed Sensing ([2]). The proposed approach combines a first step -performed at the data acquisition level- with an energy based algorithm ([3], [4]) aimed at defining a global metric on the data. The latter is then used to drive the classification algorithm. We demonstrate the power of the new technique by applying it to the detection of cellular nuclei in large, high-dimensional, hyperspectral images.
A variational approach based on level set methods popular in image segmentation is presented for learning discriminative classifiers in general feature spaces. Nonlinear, nonparametric decision boundaries are obtained...
详细信息
ISBN:
(纸本)9781424423750
A variational approach based on level set methods popular in image segmentation is presented for learning discriminative classifiers in general feature spaces. Nonlinear, nonparametric decision boundaries are obtained by minimizing an energy functional that incorporates a margin-based loss function. The class of level set contour decision boundaries is discussed in terms of the structural risk minimization principle. A variation on El feature subset selection is developed. Use of level set classifiers as base learners for boosting is discussed.
signalprocessing is multidisciplinary in nature. It provides mathematical analysis and computational operations on a wide range of signal or information types in diverse application fields that are typically classifi...
详细信息
signalprocessing is multidisciplinary in nature. It provides mathematical analysis and computational operations on a wide range of signal or information types in diverse application fields that are typically classified as different technical areas. The idea of benefiting from research methodologies and techniques across disparate but related signalprocessing technical areas has been embraced by numerous signalprocessing researchers. This summer, I had the opportunity to co-organize the Banff workshop on Multimedia, Mathematics, and machinelearning with Prof. Rabab Ward, where a group of distinguished researchers and educators worldwide were invited. Many of the invitees were pursuing research that touched on not just one but multiple signalprocessing technical areas, and thus were able to discuss common underlying principles and methods for a wide range of media signalprocessing applications, and benefit from "cross-pollination" over these fields. Several talks focused on cross-fertilization between different signalprocessing areas and these led to many interesting discussions at the workshop. Such talks included "Mobile Image Matching --- Recognition Meets Compression" by B. Girod (Stanford University), "machine Hearing (vs. machine Vision)" by D. Lyon (Google Research), and "Statistical Methods for Image, Speech, and Language processing: Achievements and Open Problems" by H. Ney (RWTH Aachen University).
We consider a likelihood framework for analyzing multivariate time series as mixtures of independent linear processes. We propose a flexible, Newton algorithm for estimating impulse response functions associated with ...
详细信息
ISBN:
(纸本)9781479911806
We consider a likelihood framework for analyzing multivariate time series as mixtures of independent linear processes. We propose a flexible, Newton algorithm for estimating impulse response functions associated with independent linear processes and an EM-based finite mixture model to handle intermittent regimes. Simulations and application to EEG are also provided.
A novel real-time acoustic feedback (RTAF) based on machinelearning to reduce the duration and to improve the progress in the rehabilitation is presented. Wearable technology (WT) has emerged as a viable means to pro...
详细信息
ISBN:
(纸本)9781538654774
A novel real-time acoustic feedback (RTAF) based on machinelearning to reduce the duration and to improve the progress in the rehabilitation is presented. Wearable technology (WT) has emerged as a viable means to provide low-cost digital healthcare and therapy course outside the medical environment like hospitals and clinics. In this paper we show that the RTAF together with WTs can offer an excellent solution to be used in rehabilitation. The method of RTAF based on machinelearning as well as a study for proving its effectiveness are presented below. The results show a faster recovery time using RTAF. The proposed RTAF shows a great potential to be used and deployed to support digital healthcare, therapy and rehabilitation.
Classical supervised learning via empirical risk (or negative log-likelihood) minimization hinges upon the assumption that the testing distribution coincides with the training distribution. This assumption can be chal...
详细信息
ISBN:
(纸本)9781728166629
Classical supervised learning via empirical risk (or negative log-likelihood) minimization hinges upon the assumption that the testing distribution coincides with the training distribution. This assumption can be challenged in modern applications of machinelearning in which learningmachines may operate at prediction time with testing data whose distribution departs from the one of the training data. We revisit the superquantile regression method by proposing a first-order optimization algorithm to minimize a superquantile-based learning objective. The proposed algorithm is based on smoothing the superquantile function by infimal convolution. Promising numerical results illustrate the interest of the approach towards safer supervised learning.
This paper presents the application of the kernel signal to noise ratio (KSNR) in the context of feature extraction to general machinelearning and signalprocessing domains. The proposed approach maximizes the signal...
详细信息
ISBN:
(纸本)9781467310260
This paper presents the application of the kernel signal to noise ratio (KSNR) in the context of feature extraction to general machinelearning and signalprocessing domains. The proposed approach maximizes the signal variance while minimizes the estimated noise variance in a reproducing kernel Hilbert space (RKHS). The KSNR can be used in any kernel method to deal with correlated (possibly non-Gaussian) noise. We illustrate the method in nonlinear regression examples, dependence estimation and causal inference, nonlinear channel equalization, and nonlinear feature extraction from high-dimensional satellite images. Results show that the proposed KSNR yields more fitted solutions and extracts more noise-free features when confronted with standard approaches.
Although optimality of sequential tests for the detection of a change in the parameter of a model has been widely discussed, the test parameter tuning is still an issue. In this communication, we propose a learning st...
详细信息
ISBN:
(纸本)0780395174
Although optimality of sequential tests for the detection of a change in the parameter of a model has been widely discussed, the test parameter tuning is still an issue. In this communication, we propose a learning strategy to set the threshold of the GLR CUSUM statistics to take a decision with a desired false alarm probability. Only data be ore the change point are required to perform the learning process. Extensive simulations are performed to assess the validity of the proposed method. The paper is concluded by opening the path to a new approach to multi-modal feature based event detection for video parsing.
暂无评论