scholarly journals Robust high-dimensional memory-augmented neural networks

2021 ◽  
Vol 12 (1) ◽  
Author(s):  
Geethan Karunaratne ◽  
Manuel Schmuck ◽  
Manuel Le Gallo ◽  
Giovanni Cherubini ◽  
Luca Benini ◽  
...  
Keyword(s):  
Neural Networks ◽  
Computer Architecture ◽  
Explicit Memory ◽  
High Dimensional ◽  
Training Procedure ◽  
Von Neumann ◽  
Data Memory ◽  
Individual Memory ◽  
Classification Tasks ◽  
Computational Memory

AbstractTraditional neural networks require enormous amounts of data to build their complex mappings during a slow training procedure that hinders their abilities for relearning and adapting to new data. Memory-augmented neural networks enhance neural networks with an explicit memory to overcome these issues. Access to this explicit memory, however, occurs via soft read and write operations involving every individual memory entry, resulting in a bottleneck when implemented using the conventional von Neumann computer architecture. To overcome this bottleneck, we propose a robust architecture that employs a computational memory unit as the explicit memory performing analog in-memory computation on high-dimensional (HD) vectors, while closely matching 32-bit software-equivalent accuracy. This is achieved by a content-based attention mechanism that represents unrelated items in the computational memory with uncorrelated HD vectors, whose real-valued components can be readily approximated by binary, or bipolar components. Experimental results demonstrate the efficacy of our approach on few-shot image classification tasks on the Omniglot dataset using more than 256,000 phase-change memory devices. Our approach effectively merges the richness of deep neural network representations with HD computing that paves the way for robust vector-symbolic manipulations applicable in reasoning, fusion, and compression.

scholarly journals Evaluation of Mixed Deep Neural Networks for Reverberant Speech Enhancement

Biomimetics ◽  
2019 ◽  
Vol 5 (1) ◽  
pp. 1 ◽  
Author(s):  
Michelle Gutiérrez-Muñoz ◽  
Astryd González-Salazar ◽  
Marvin Coto-Jiménez
Keyword(s):  
Neural Networks ◽  
Short Term Memory ◽  
Computational Cost ◽  
Real Life ◽  
Fixed Number ◽  
Training Procedure ◽  
Statistical Validation ◽  
Significant Drop ◽  
Training Time ◽  
Important Solution

Speech signals are degraded in real-life environments, as a product of background noise or other factors. The processing of such signals for voice recognition and voice analysis systems presents important challenges. One of the conditions that make adverse quality difficult to handle in those systems is reverberation, produced by sound wave reflections that travel from the source to the microphone in multiple directions. To enhance signals in such adverse conditions, several deep learning-based methods have been proposed and proven to be effective. Recently, recurrent neural networks, especially those with long short-term memory (LSTM), have presented surprising results in tasks related to time-dependent processing of signals, such as speech. One of the most challenging aspects of LSTM networks is the high computational cost of the training procedure, which has limited extended experimentation in several cases. In this work, we present a proposal to evaluate the hybrid models of neural networks to learn different reverberation conditions without any previous information. The results show that some combinations of LSTM and perceptron layers produce good results in comparison to those from pure LSTM networks, given a fixed number of layers. The evaluation was made based on quality measurements of the signal’s spectrum, the training time of the networks, and statistical validation of results. In total, 120 artificial neural networks of eight different types were trained and compared. The results help to affirm the fact that hybrid networks represent an important solution for speech signal enhancement, given that reduction in training time is on the order of 30%, in processes that can normally take several days or weeks, depending on the amount of data. The results also present advantages in efficiency, but without a significant drop in quality.

Neural networks trained with high-dimensional functions approximation data in high-dimensional space

2021 ◽  
pp. 1-12
Author(s):  
Jian Zheng ◽  
Jianfeng Wang ◽  
Yanping Chen ◽  
Shuping Chen ◽  
Jingjin Chen ◽  
...  
Keyword(s):  
Neural Networks ◽  
Dimensional Space ◽  
Data Distribution ◽  
High Dimensional ◽  
Sufficient Information ◽  
Sufficient Data ◽  
High Dimensional Space ◽  
Positive Effects ◽  
The Neural Networks ◽  
Using Data

Neural networks can approximate data because of owning many compact non-linear layers. In high-dimensional space, due to the curse of dimensionality, data distribution becomes sparse, causing that it is difficulty to provide sufficient information. Hence, the task becomes even harder if neural networks approximate data in high-dimensional space. To address this issue, according to the Lipschitz condition, the two deviations, i.e., the deviation of the neural networks trained using high-dimensional functions, and the deviation of high-dimensional functions approximation data, are derived. This purpose of doing this is to improve the ability of approximation high-dimensional space using neural networks. Experimental results show that the neural networks trained using high-dimensional functions outperforms that of using data in the capability of approximation data in high-dimensional space. We find that the neural networks trained using high-dimensional functions more suitable for high-dimensional space than that of using data, so that there is no need to retain sufficient data for neural networks training. Our findings suggests that in high-dimensional space, by tuning hidden layers of neural networks, this is hard to have substantial positive effects on improving precision of approximation data.

scholarly journals Automatic Evaluation of the Lung Condition of COVID-19 Patients Using X-ray Images and Convolutional Neural Networks

2021 ◽  
Vol 11 (1) ◽  
pp. 28
Author(s):  
Ivan Lorencin ◽  
Sandi Baressi Šegota ◽  
Nikola Anđelić ◽  
Anđela Blagojević ◽  
Tijana Šušteršić ◽  
...  
Keyword(s):  
Neural Networks ◽  
Convolutional Neural Networks ◽  
Clinical Picture ◽  
Data Augmentation ◽  
Lower Amount ◽  
Training Procedure ◽  
X Ray ◽  
Severe Clinical Picture ◽  
Lung Condition

COVID-19 represents one of the greatest challenges in modern history. Its impact is most noticeable in the health care system, mostly due to the accelerated and increased influx of patients with a more severe clinical picture. These facts are increasing the pressure on health systems. For this reason, the aim is to automate the process of diagnosis and treatment. The research presented in this article conducted an examination of the possibility of classifying the clinical picture of a patient using X-ray images and convolutional neural networks. The research was conducted on the dataset of 185 images that consists of four classes. Due to a lower amount of images, a data augmentation procedure was performed. In order to define the CNN architecture with highest classification performances, multiple CNNs were designed. Results show that the best classification performances can be achieved if ResNet152 is used. This CNN has achieved AUCmacro¯ and AUCmicro¯ up to 0.94, suggesting the possibility of applying CNN to the classification of the clinical picture of COVID-19 patients using an X-ray image of the lungs. When higher layers are frozen during the training procedure, higher AUCmacro¯ and AUCmicro¯ values are achieved. If ResNet152 is utilized, AUCmacro¯ and AUCmicro¯ values up to 0.96 are achieved if all layers except the last 12 are frozen during the training procedure.

Recurrent Neural Networks for Edge Intelligence

2021 ◽  
Vol 54 (4) ◽  
pp. 1-38
Author(s):  
Varsha S. Lalapura ◽  
J. Amudha ◽  
Hariramn Selvamuruga Satheesh
Keyword(s):  
Neural Networks ◽  
Recurrent Neural Networks ◽  
Model Performance ◽  
Network Architectures ◽  
Training Procedure ◽  
Resource Constrained ◽  
Compression Technique ◽  
Creative Art ◽  
Comprehensive Survey ◽  
Intelligent Models

Recurrent Neural Networks are ubiquitous and pervasive in many artificial intelligence applications such as speech recognition, predictive healthcare, creative art, and so on. Although they provide accurate superior solutions, they pose a massive challenge “training havoc.” Current expansion of IoT demands intelligent models to be deployed at the edge. This is precisely to handle increasing model sizes and complex network architectures. Design efforts to meet these for greater performance have had inverse effects on portability on edge devices with real-time constraints of memory, latency, and energy. This article provides a detailed insight into various compression techniques widely disseminated in the deep learning regime. They have become key in mapping powerful RNNs onto resource-constrained devices. While compression of RNNs is the main focus of the survey, it also highlights challenges encountered while training. The training procedure directly influences model performance and compression alongside. Recent advancements to overcome the training challenges with their strengths and drawbacks are discussed. In short, the survey covers the three-step process, namely, architecture selection, efficient training process, and suitable compression technique applicable to a resource-constrained environment. It is thus one of the comprehensive survey guides a developer can adapt for a time-series problem context and an RNN solution for the edge.

scholarly journals High-order compact LOD methods for solving high-dimensional advection equations

2021 ◽  
Vol 40 (3) ◽  
Author(s):  
Bo Hou ◽  
Yongbin Ge
Keyword(s):  
Truncation Error ◽  
Three Dimensional ◽  
Taylor Series Expansion ◽  
High Order ◽  
High Dimensional ◽  
Analysis Method ◽  
Order Accuracy ◽  
Von Neumann ◽  
One Dimensional ◽  
Von Neumann Analysis

AbstractIn this paper, by using the local one-dimensional (LOD) method, Taylor series expansion and correction for the third derivatives in the truncation error remainder, two high-order compact LOD schemes are established for solving the two- and three- dimensional advection equations, respectively. They have the fourth-order accuracy in both time and space. By the von Neumann analysis method, it shows that the two schemes are unconditionally stable. Besides, the consistency and convergence of them are also proved. Finally, numerical experiments are given to confirm the accuracy and efficiency of the present schemes.

scholarly journals Efficient approximation of solutions of parametric linear transport equations by ReLU DNNs

2021 ◽  
Vol 47 (1) ◽  
Author(s):  
Fabian Laakmann ◽  
Philipp Petersen
Keyword(s):  
Neural Networks ◽  
Deep Neural Networks ◽  
Initial Conditions ◽  
Activation Function ◽  
Transport Equations ◽  
High Dimensional ◽  
Linear Transport ◽  
Approximation Rates ◽  
Curse Of Dimension ◽  
Efficient Approximation

AbstractWe demonstrate that deep neural networks with the ReLU activation function can efficiently approximate the solutions of various types of parametric linear transport equations. For non-smooth initial conditions, the solutions of these PDEs are high-dimensional and non-smooth. Therefore, approximation of these functions suffers from a curse of dimension. We demonstrate that through their inherent compositionality deep neural networks can resolve the characteristic flow underlying the transport equations and thereby allow approximation rates independent of the parameter dimension.

scholarly journals A Neural Network Decomposition Algorithm for Mapping on Crossbar-Based Computing Systems

Electronics ◽  
2020 ◽  
Vol 9 (9) ◽  
pp. 1526 ◽  
Author(s):  
Choongmin Kim ◽  
Jacob A. Abraham ◽  
Woochul Kang ◽  
Jaeyong Chung
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Predictive Performance ◽  
Fixed Number ◽  
Decomposition Algorithm ◽  
Computing Systems ◽  
Network Decomposition ◽  
Von Neumann ◽  
Analog Computing ◽  
The Neural Network

Crossbar-based neuromorphic computing to accelerate neural networks is a popular alternative to conventional von Neumann computing systems. It is also referred as processing-in-memory and in-situ analog computing. The crossbars have a fixed number of synapses per neuron and it is necessary to decompose neurons to map networks onto the crossbars. This paper proposes the k-spare decomposition algorithm that can trade off the predictive performance against the neuron usage during the mapping. The proposed algorithm performs a two-level hierarchical decomposition. In the first global decomposition, it decomposes the neural network such that each crossbar has k spare neurons. These neurons are used to improve the accuracy of the partially mapped network in the subsequent local decomposition. Our experimental results using modern convolutional neural networks show that the proposed method can improve the accuracy substantially within about 10% extra neurons.

Sign in / Sign up

Export Citation Format

Share Document