The Infona portal uses cookies, i.e. strings of text saved by a browser on the user's device. The portal can access those files and use them to remember the user's data, such as their chosen settings (screen view, interface language, etc.), or their login data. By using the Infona portal the user accepts automatic saving and using this information for portal operation purposes. More information on the subject can be found in the Privacy Policy and Terms of Service. By closing this window the user confirms that they have read the information on cookie usage, and they accept the privacy policy and the way cookies are used by the portal. You can change the cookie settings in your browser.
Recurrent neural network has been widely used as auto-regressive model for time series. The most commonly used training method for recurrent neural network is back propagation. However, recurrent neural networks trained with back propagation can get trapped at local minima and saddle points. In these cases, auto-regressive models cannot effectively model time series patterns. In order to address these...
In this paper, we present a method for stereo super-resolution which employs a deep network. The network is trained using the residual image so as to obtain a high resolution image from two, low resolution views. Our network is comprised by two deep sub-nets which share, at their output, a single convolutional layer. This last layer in the network delivers an estimate of the residual image which is...
Joint sparse representation (JSR) models have been widely applied into the field of hyperspectral image (HSI) classification. However, most of JSR-based models adopt the Frobenius norm to measure the reconstruction error, which ignores the structural information of the small patch. In this paper, we propose a nuclear-norm joint sparse representation (NuJSR) model for hyperspectral image classification...
Most previous algorithms for the recognition of Action Units (AUs) were trained on a small number of sample images. This was due to the limited amount of labeled data available at the time. This meant that data-hungry deep neural networks, which have shown their potential in other computer vision problems, could not be successfully trained to detect AUs. A recent publicly available database with close...
There have been significant progresses in single image super-resolution (SR) using deep convolutional neural network. In this paper, we propose a modified deep convolutional neural network model incorporated with image texture priors for single image SR. The model consist of a particular feature extraction layer followed by image reconstruction process, aiming to centralize on the image texture information...
Deep embeddings answer one simple question: How similar are two images? Learning these embeddings is the bedrock of verification, zero-shot learning, and visual search. The most prominent approaches optimize a deep convolutional network with a suitable loss function, such as contrastive loss or triplet loss. While a rich line of work focuses solely on the loss functions, we show in this paper that...
Concept of computing and communication convergence has many years of history — as strategic concept it started about 1977 and yet have many definitions. The idea behind convergence concept mostly related to getting closer computing and telecommunication common technologies, services and service provider's business models. In this academic position paper some less significant obstacles of influence-reflection...
To solve the problem of low recognition rate which is the existing identification methods of partial discharge faults, a new method was designed with wavelet, singular value and improved particle swarm algorithm to optimize the BP neural network. First, using continuous wavelet and singular value decomposition to get the signal characteristic value; then combined with the significance of inertia weight...
We address the problem of distance metric learning (DML), defined as learning a distance consistent with a notion of semantic similarity. Traditionally, for this problem supervision is expressed in the form of sets of points that follow an ordinal relationship – an anchor point x is similar to a set of positive points Y , and dissimilar to a set of negative points Z, and a loss defined over these...
This paper carries out a large dimensional analysis of the standard regularized quadratic discriminant analysis (QDA) classifier designed on the assumption that data arise from a Gaussian mixture model. The analysis relies on fundamental results from random matrix theory (RMT) when both the number of features and the cardinality of the training data within each class grow large at the same pace. Under...
This paper proposes a regrouping particle swarm optimization-based neural network (RegPSONN) for rolling bearing fault diagnosis. The proposed method applied neural network for rolling bearing conditions classification, and regrouping particle swarm optimization (RegPSO) is utilized for network training, and ten time-domain feature parameters are selected to establish the input vector. To evaluate...
Dancers are human Expressive Motion experts and could theoretically help robots communicate their state to people, e.g., rushed, confused, curious. The problem is twofold: first, dancers are trained in human-motion whereas many robots are non-anthropomorphic, and second, most dancers are not programmers. This is where the present interface is useful: the robot demos a batch of motions, in person,...
Parallel and distributed processing is employed to accelerate training for many deep-learning applications with large models and inputs. As it reduces synchronization and communication overhead by tolerating stale gradient updates, asynchronous stochastic gradient descent (ASGD), derived from stochastic gradient descent (SGD), is widely used. Recent theoretical analyses show ASGD converges with linear...
Nowadays, applications based on digits recognition and characters recognition have become much more reliable thanks to the rapid development of the DNN(deep neural network) architecture and constantly increasing the efficiency to the computing resources. A lot of methods have been proposed to improve the performance of DNNs, such as the ReLU (Rectified Linear Unit) which is a widely used alternative...
Accelerating the inference of a trained DNN is a well studied subject. In this paper we switch the focus to the training of DNNs. The training phase is compute intensive, demands complicated data communication, and contains multiple levels of data dependencies and parallelism. This paper presents an algorithm/architecture space exploration of efficient accelerators to achieve better network convergence...
Convolutional dictionary learning (CDL) has great potential to “learn” rich sparse representations from training datasets, by training translation-invariant filters. However, the performance of applying learned filters from CDL to inverse problems has not yet been fully maximized because training data preprocessing in training stage is not fully compensated in testing stage. We propose CDL using Adaptive...
In this paper, we focus on promoting multi-label learning task with ensemble learning. Compared to traditional single algorithm methods, it has been recognized that ensemble methods could achieve much better performance than each constituent learned model, especially under the conditional independence of different classifiers. Existing multi-label ensemble algorithms mainly focus on creating diverse...
This paper presents a novel nonlinear adaptive filter method, namely, Hammerstein adaptive filter with single feedback under minimum mean square error (HAF-SF-MMSE). A single delayed output is incorporated into the estimation of the current output based on minimum mean square error criterion, and therefore the history information of output is considered. Moreover, hybrid learning rates and adaptive...
Independent Component Analysis (ICA) is a dimensionality reduction technique that can boost efficiency of machine learning models that deal with probability density functions, e.g. Bayesian neural networks. Algorithms that implement adaptive ICA converge slower than their nonadaptive counterparts, however, they are capable of tracking changes in underlying distributions of input features. This intrinsically...
Vehicle logo recognition is an important part of vehicle identification in intelligent transportation systems. State-of-the-art vehicle logo recognition approaches typically consider training models on large datasets. However, there might only be a small training dataset to start with and more images can be obtained during the real-time applications. This paper proposes an online image recognition...
Set the date range to filter the displayed results. You can set a starting date, ending date or both. You can enter the dates manually or choose them from the calendar.