scholarly journals A Study on Effectiveness of Deep Neural Networks for Speech Signal Enhancement in Comparison with Wiener Filtering Technique

2021 ◽  
Author(s):  
Vijaya Kumar Padarti ◽  
Gnana Sai Polavarapu ◽  
Madhurima Madiraju ◽  
Naga Sai Nuthalapati ◽  
Vinay Babu Thota ◽  
...  
Keyword(s):  
Neural Network ◽  
Speech Signal ◽  
Deep Neural Networks ◽  
Network Models ◽  
Signal Enhancement ◽  
Wiener Filtering ◽  
Quality Enhancement ◽  
Neural Network Models ◽  
Filtering Technique ◽  
Speech Denoising

We have compared two Neural network models with Wiener filtering technique for Speech signal enhancement. Our paper intends to suggest the best method suitable for speech denoising and quality enhancement. We have utilized MATLAB software with most advanced toolboxes for building the models. For comparing our models, we computed PSNR and SNR values.

scholarly journals A Study on Effectiveness of Deep Neural Networks for Speech Signal Enhancement in Comparison with Wiener Filtering Technique

2021 ◽  
Author(s):  
Vijaya Kumar Padarti ◽  
Gnana Sai Polavarapu ◽  
Madhurima Madiraju ◽  
Naga Sai Nuthalapati ◽  
Vinay Babu Thota ◽  
...  
Keyword(s):  
Neural Network ◽  
Speech Signal ◽  
Deep Neural Networks ◽  
Network Models ◽  
Signal Enhancement ◽  
Wiener Filtering ◽  
Quality Enhancement ◽  
Neural Network Models ◽  
Filtering Technique ◽  
Speech Denoising

We have compared two Neural network models with Wiener filtering technique for Speech signal enhancement. Our paper intends to suggest the best method suitable for speech denoising and quality enhancement. We have utilized MATLAB software with most advanced toolboxes for building the models. For comparing our models, we computed PSNR and SNR values.

Explainable Deep Learning Reproduces a ‘Professional Eye’ on the Diagnosis of Internal Disorders in Persimmon Fruit

2020 ◽  
Vol 61 (11) ◽  
pp. 1967-1973
Author(s):  
Takashi Akagi ◽  
Masanori Onishi ◽  
Kanae Masuda ◽  
Ryohei Kuroki ◽  
Kohei Baba ◽  
...  
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Deep Neural Network ◽  
Deep Neural Networks ◽  
Binary Classification ◽  
Network Models ◽  
Plant Biology ◽  
Neural Network Models ◽  
Persimmon Fruit

Abstract Recent rapid progress in deep neural network techniques has allowed recognition and classification of various objects, often exceeding the performance of the human eye. In plant biology and crop sciences, some deep neural network frameworks have been applied mainly for effective and rapid phenotyping. In this study, beyond simple optimizations of phenotyping, we propose an application of deep neural networks to make an image-based internal disorder diagnosis that is hard even for experts, and to visualize the reasons behind each diagnosis to provide biological interpretations. Here, we exemplified classification of calyx-end cracking in persimmon fruit by using five convolutional neural network models with various layer structures and examined potential analytical options involved in the diagnostic qualities. With 3,173 visible RGB images from the fruit apex side, the neural networks successfully made the binary classification of each degree of disorder, with up to 90% accuracy. Furthermore, feature visualizations, such as Grad-CAM and LRP, visualize the regions of the image that contribute to the diagnosis. They suggest that specific patterns of color unevenness, such as in the fruit peripheral area, can be indexes of calyx-end cracking. These results not only provided novel insights into indexes of fruit internal disorders but also proposed the potential applicability of deep neural networks in plant biology.

scholarly journals A Novel Low-Bit Quantization Strategy for Compressing Deep Neural Networks

2020 ◽  
Vol 2020 ◽  
pp. 1-7
Author(s):  
Xin Long ◽  
XiangRong Zeng ◽  
Zongcheng Ben ◽  
Dianle Zhou ◽  
Maojun Zhang
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Deep Neural Networks ◽  
Computational Cost ◽  
Network Models ◽  
Memory Consumption ◽  
Neural Network Models ◽  
The Neural Network ◽  
The Neural Networks ◽  
Novel Strategy

The increase in sophistication of neural network models in recent years has exponentially expanded memory consumption and computational cost, thereby hindering their applications on ASIC, FPGA, and other mobile devices. Therefore, compressing and accelerating the neural networks are necessary. In this study, we introduce a novel strategy to train low-bit networks with weights and activations quantized by several bits and address two corresponding fundamental issues. One is to approximate activations through low-bit discretization for decreasing network computational cost and dot-product memory. The other is to specify weight quantization and update mechanism for discrete weights to avoid gradient mismatch. With quantized low-bit weights and activations, the costly full-precision operation will be replaced by shift operation. We evaluate the proposed method on common datasets, and results show that this method can dramatically compress the neural network with slight accuracy loss.

scholarly journals Surrogate Object Detection Explainer (SODEx) with YOLOv4 and LIME

2021 ◽  
Vol 3 (3) ◽  
pp. 662-671
Author(s):  
Jonas Herskind Sejr ◽  
Peter Schneider-Kamp ◽  
Naeem Ayoub
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Object Detection ◽  
Deep Neural Networks ◽  
Empirical Evaluation ◽  
Network Models ◽  
Detection Algorithm ◽  
Neural Network Models ◽  
The Neural Network

Due to impressive performance, deep neural networks for object detection in images have become a prevalent choice. Given the complexity of the neural network models used, users of these algorithms are typically given no hint as to how the objects were found. It remains, for example, unclear whether an object is detected based on what it looks like or based on the context in which it is located. We have developed an algorithm, Surrogate Object Detection Explainer (SODEx), that can explain any object detection algorithm using any classification explainer. We evaluate SODEx qualitatively and quantitatively by detecting objects in the COCO dataset with YOLOv4 and explaining these detections with LIME. This empirical evaluation does not only demonstrate the value of explainable object detection, it also provides valuable insights into how YOLOv4 detects objects.

scholarly journals Detecting failure modes in image reconstructions with interval neural network uncertainty

2021 ◽  
Author(s):  
Luis Oala ◽  
Cosmas Heiß ◽  
Jan Macdonald ◽  
Maximilian März ◽  
Gitta Kutyniok ◽  
...  
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Image Reconstruction ◽  
Deep Neural Networks ◽  
Failure Modes ◽  
Network Models ◽  
Quantitative Detection ◽  
Neural Network Models ◽  
Post Hoc ◽  
Interval Neural Network

Abstract Purpose The quantitative detection of failure modes is important for making deep neural networks reliable and usable at scale. We consider three examples for common failure modes in image reconstruction and demonstrate the potential of uncertainty quantification as a fine-grained alarm system. Methods We propose a deterministic, modular and lightweight approach called Interval Neural Network (INN) that produces fast and easy to interpret uncertainty scores for deep neural networks. Importantly, INNs can be constructed post hoc for already trained prediction networks. We compare it against state-of-the-art baseline methods (MCDrop, ProbOut). Results We demonstrate on controlled, synthetic inverse problems the capacity of INNs to capture uncertainty due to noise as well as directional error information. On a real-world inverse problem with human CT scans, we can show that INNs produce uncertainty scores which improve the detection of all considered failure modes compared to the baseline methods. Conclusion Interval Neural Networks offer a promising tool to expose weaknesses of deep image reconstruction models and ultimately make them more reliable. The fact that they can be applied post hoc to equip already trained deep neural network models with uncertainty scores makes them particularly interesting for deployment.

Building on prior knowledge without building it in

2017 ◽  
Vol 40 ◽  
Author(s):  
Steven S. Hansen ◽  
Andrew K. Lampinen ◽  
Gaurav Suri ◽  
James L. McClelland
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Prior Knowledge ◽  
Deep Neural Network ◽  
Deep Neural Networks ◽  
Network Models ◽  
Causal Models ◽  
Neural Network Models ◽  
Start Up ◽  
The Many

AbstractLake et al. propose that people rely on “start-up software,” “causal models,” and “intuitive theories” built using compositional representations to learn new tasks more efficiently than some deep neural network models. We highlight the many drawbacks of a commitment to compositional representations and describe our continuing effort to explore how the ability to build on prior knowledge and to learn new tasks efficiently could arise through learning in deep neural networks.

scholarly journals Determining the Age of the Author of the Text Based on Deep Neural Network Models

Information ◽  
2020 ◽  
Vol 11 (12) ◽  
pp. 589 ◽  
Author(s):  
Aleksandr Sergeevich Romanov ◽  
Anna Vladimirovna Kurtukova ◽  
Artem Alexandrovich Sobolev ◽  
Alexander Alexandrovich Shelupanov ◽  
Anastasia Mikhailovna Fedotova
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Deep Neural Network ◽  
Deep Neural Networks ◽  
Network Models ◽  
Russian Language ◽  
Neural Network Models ◽  
Inaccurate Data ◽  
Language Text

This paper is devoted to solving the problem of determining the age of the author of the text based on models of deep neural networks. The article presents an analysis of methods for determining the age of the author of a text and approaches to determining the age of a user by a photo. This could be a solution to the problem of inaccurate data for training by filtering out incorrect user-specified age data. A detailed description of the author’s technique based on deep neural network models and the interpretation of the results is also presented. The study found that the proposed technique achieved 82% accuracy in determining the age of the author from Russian-language text, which makes it competitive in comparison with approaches for other languages.

scholarly journals Performance analysis of the TensorFlow library with different optimisation algorithms

2021 ◽  
Vol 21 ◽  
pp. 330-335
Author(s):  
Maciej Wadas ◽  
Jakub Smołka
Keyword(s):  
Neural Network ◽  
Performance Analysis ◽  
Network Model ◽  
Neural Network Model ◽  
Deep Neural Networks ◽  
Network Models ◽  
Neural Network Models ◽  
Training Time ◽  
The Neural Network ◽  
Training Efficiency

This paper presents the results of performance analysis of the Tensorflow library used in machine learning and deep neural networks. The analysis focuses on comparing the parameters obtained when training the neural network model for optimization algorithms: Adam, Nadam, AdaMax, AdaDelta, AdaGrad. Special attention has been paid to the differences between the training efficiency on tasks using microprocessor and graphics card. For the study, neural network models were created in order to recognise Polish handwritten characters. The results obtained showed that the most efficient algorithm is AdaMax, while the computer component used during the research only affects the training time of the neural network model used.

scholarly journals CDeep3M-Preview: Online segmentation using the deep neural network model zoo

2020 ◽  
Author(s):  
Matthias G Haberl ◽  
Willy Wong ◽  
Sean Penticoff ◽  
Jihyeon Je ◽  
Matthew Madany ◽  
...  
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Deep Neural Networks ◽  
Network Models ◽  
Large Datasets ◽  
Neural Network Models ◽  
Online Tool ◽  
Trained Neural Network ◽  
Cloud Applications ◽  
Microscopy Images

AbstractSharing deep neural networks and testing the performance of trained networks typically involves a major initial commitment towards one algorithm, before knowing how the network will perform on a different dataset. Here we release a free online tool, CDeep3M-Preview, that allows end-users to rapidly test the performance of any of the pre-trained neural network models hosted on the CIL-CDeep3M modelzoo. This feature makes part of a set of complementary strategies we employ to facilitate sharing, increase reproducibility and enable quicker insights into biology. Namely we: (1) provide CDeep3M deep learning image segmentation software through cloud applications (Colab and AWS) and containerized installations (Docker and Singularity) (2) co-hosting trained deep neural networks with the relevant microscopy images and (3) providing a CDeep3M-Preview feature, enabling quick tests of trained networks on user provided test data or any of the publicly hosted large datasets. The CDeep3M-modelzoo and the cellimagelibrary.org are open for contributions of both, trained models as well as image datasets by the community and all services are free of charge.

scholarly journals Dataset for Eye-Tracking Tasks

2020 ◽  
Author(s):  
Ildar Rakhmatulin
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Eye Tracking ◽  
Deep Neural Networks ◽  
Network Models ◽  
Neural Network Models ◽  
Data Set ◽  
Tracking Tasks ◽  
Deep Networks ◽  
Using Data

In recent years many different deep neural networks were developed, but due to a large number of layers in deep networks, their training requires a long time and a large number of datasets. Today is popular to use trained deep neural networks for various tasks, even for simple ones in which such deep networks are not required. The well-known deep networks such as YoloV3, SSD, etc. are intended for tracking and monitoring various objects, therefore their weights are heavy and the overall accuracy for a specific task is low. Eye-tracking tasks need to detect only one object - an iris in a given area. Therefore, it is logical to use a neural network only for this task. But the problem is the lack of suitable datasets for training the model. In the manuscript, we presented a dataset that is suitable for training custom models of convolutional neural networks for eye-tracking tasks. Using data set data, each user can independently pre-train the convolutional neural network models for eye-tracking tasks. This dataset contains annotated 10,000 eye images in an extension of 416 by 416 pixels. The table with annotation information shows the coordinates and radius of the eye for each image. This manuscript can be considered as a guide for the preparation of datasets for eye-tracking devices.

Sign in / Sign up

Export Citation Format

Share Document