Invertible Grayscale with Sparsity Enforcing Priors

2021 ◽  
Vol 17 (3) ◽  
pp. 1-17
Author(s):  
Yong Du ◽  
Yangyang Xu ◽  
Taizhong Ye ◽  
Qiang Wen ◽  
Chufeng Xiao ◽  
...  
Keyword(s):  
Neural Network ◽  
Deep Neural Network ◽  
Storage Capacity ◽  
Color Image ◽  
Grayscale Image ◽  
Storage Space ◽  
Color Information ◽  
Compression Process ◽  
The Arts ◽  
Residual Image

Color dimensionality reduction is believed as a non-invertible process, as re-colorization results in perceptually noticeable and unrecoverable distortion. In this article, we propose to convert a color image into a grayscale image that can fully recover its original colors, and more importantly, the encoded information is discriminative and sparse, which saves storage capacity. Particularly, we design an invertible deep neural network for color encoding and decoding purposes. This network learns to generate a residual image that encodes color information, and it is then combined with a base grayscale image for color recovering. In this way, the non-differentiable compression process (e.g., JPEG) of the base grayscale image can be integrated into the network in an end-to-end manner. To further reduce the size of the residual image, we present a specific layer to enhance Sparsity Enforcing Priors (SEP), thus leading to negligible storage space. The proposed method allows color embedding on a sparse residual image while keeping a high, 35dB PSNR on average. Extensive experiments demonstrate that the proposed method outperforms state-of-the-arts in terms of image quality and tolerability to compression.

Acoustic Scenery Recognition Using CWT and Deep Neural Network

2021 ◽  
Author(s):  
Francisco Mondragon ◽  
Jonathan Jimenez ◽  
Mariko Nakano ◽  
Toru Nakashika ◽  
Hector Perez-Meana
Keyword(s):  
Neural Network ◽  
Deep Neural Network ◽  
Color Image ◽  
Classification Performance ◽  
Science And Engineering ◽  
Time Frequency ◽  
Frequency Representation ◽  
Public Data ◽  
Recognition Systems ◽  
Color Map

The development of acoustic scenes recognition systems has been a topic of extensive research due to its applications in several fields of science and engineering. This paper proposes an environmental system in which firstly a time-frequency representation is obtained using the Continuous Wavelet Transform (CWT). The time frequency representation is then represented as a color image using the Viridis color map, which is then inserted into a Deep Neural Network (DNN) to carry out the classification task. Evaluation results using several public data bases show that proposed scheme provides a classification performance better than the performance provided by other previously proposed schemes.

scholarly journals Learning-Based Colorization of Grayscale Aerial Images Using Random Forest Regression

2018 ◽  
Vol 8 (8) ◽  
pp. 1269 ◽  
Author(s):  
Dae Seo ◽  
Yong Kim ◽  
Yang Eo ◽  
Wan Park
Keyword(s):  
Random Forest ◽  
Color Image ◽  
Color Space ◽  
Elevation Angle ◽  
Aerial Images ◽  
Aerial Image ◽  
Grayscale Image ◽  
Color Information ◽  
Image Colorization ◽  
Lab Color Space

Image colorization assigns colors to a grayscale image, which is an important yet difficult image-processing task encountered in various applications. In particular, grayscale aerial image colorization is a poorly posed problem that is affected by the sun elevation angle, seasons, sensor parameters, etc. Furthermore, since different colors may have the same intensity, it is difficult to solve this problem using traditional methods. This study proposes a novel method for the colorization of grayscale aerial images using random forest (RF) regression. The algorithm uses one grayscale image for input and one-color image for reference, both of which have similar seasonal features at the same location. The reference color image is then converted from the Red-Green-Blue (RGB) color space to the CIE L*a*b (Lab) color space in which the luminance is used to extract training pixels; this is done by performing change detection with the input grayscale image, and color information is used to establish color relationships. The proposed method directly establishes color relationships between features of the input grayscale image and color information of the reference color image based on the corresponding training pixels. The experimental results show that the proposed method outperforms several state-of-the-art algorithms in terms of both visual inspection and quantitative evaluation.

scholarly journals Abnormal electricity detection with hybrid deep neural network model

2018 ◽  
Vol 189 ◽  
pp. 03001
Author(s):  
Jie Liu ◽  
Xiang Cao ◽  
Diangang Wang ◽  
Kejia Pan ◽  
Cheng Zhang ◽  
...  
Keyword(s):  
Neural Network ◽  
Power Consumption ◽  
Large Scale ◽  
Deep Neural Network ◽  
Electricity Consumption ◽  
The State ◽  
Detection Model ◽  
Gradient Model ◽  
The Arts ◽  
The Difference

This paper tackles a new challenge in abnormal electricity detection: how to promptly detect stealing electricity behavior by a large-scale data from power users. Proposed scheme firstly forms power consumption gradient model by extracting daily trend indicators of electricity consumption, which can exactly reflect the short-term power consumption trend for each user. Furthermore, we design the line-losing model by analyzing the difference between power supplying and actual power consumption. Finally, a hybrid deep neural network detection model is built by combining with the power consumption gradient model and the line-losing model, which can quickly pin down to the abnormal electricity users. Comprehensive experiments are implemented by large-scale user samples from the State Grid Corporation and Tensorflow framework. Extensive results show that comparing with the state-of-the-arts, proposed scheme has a superior detection performance, and therefore is believed to be able to give a better guidance to abnormal electricity detection.

An approach to remove duplication records in healthcare dataset based on Mimic Deep Neural Network (MDNN) and Chaotic Whale Optimization (CWO)

2021 ◽  
pp. 1063293X2199201
Author(s):  
Anto Praveena M.D. ◽  
Bharathi B
Keyword(s):  
Neural Network ◽  
Data Storage ◽  
Data Privacy ◽  
Deep Neural Network ◽  
Cost Effective ◽  
Healthcare Sector ◽  
Storage Space ◽  
Model And Simulation ◽  
Whale Optimization ◽  
Space Data

Duplication of data in an application will become an expensive factor. These replication of data need to be checked and if it is needed it has to be removed from the dataset as it occupies huge volume of data in the storage space. The cloud is the main source of data storage and all organizations are already started to move their dataset into the cloud since it is cost effective, storage space, data security and data Privacy. In the healthcare sector, storing the duplicated records leads to wrong prediction. Also uploading same files by many users, data storage demand will be occurred. To address those issues, this paper proposes an Optimal Removal of Deduplication (ORD) in heart disease data using hybrid trust based neural network algorithm. In ORD scheme, the Chaotic Whale Optimization (CWO) algorithm is used for trust computation of data using multiple decision metrics. The computed trust values and the nature of the data’s are sequentially applied to the training process by the Mimic Deep Neural Network (MDNN). It classify the data is a duplicate or not. Hence the duplicates files are identified and they were removed from the data storage. Finally, the simulation evaluates to examine the proposed MDNN based model and simulation results show the effectiveness of ORD scheme in terms of data duplication removal. From the simulation result it is found that the model’s accuracy, sensitivity and specificity was good.

A Proposed Grayscale Face Image Colorization System using Particle Swarm Optimization

2017 ◽  
Vol 1 (1) ◽  
pp. 72-89 ◽  
Author(s):  
Abul Hasnat ◽  
Santanu Halder ◽  
Debotosh Bhattacharjee ◽  
Mita Nasipuri
Keyword(s):  
Real Time ◽  
Color Image ◽  
Turing Test ◽  
Face Image ◽  
Reference Image ◽  
Grayscale Image ◽  
Color Information ◽  
Image Colorization ◽  
Face Images ◽  
Real Time Applications

The proposed work is a novel grayscale face image colorization approach using a reference color face image. It takes a reference color image which presumably contains semantically similar color information for the query grayscale image and colorizes the grayscale face image with the help of the reference image. In this novel patch based colorization, the system searches a suitable patch on reference color image for each patch of grayscale image to colorize. Exhaustive patch search in reference color image takes much time resulting slow colorization process applicable for real time applications. So PSO is used to reduce the patch searching time for faster colorization process applicable in real time applications. The proposed method is successfully applied on 150 male and female face images of FRAV2D database. “Colorization Turing test” was conducted asking human subject to choose the image(close to the original color image) between colorized image using proposed algorithm and recent methods and in most of the cases colorized images using the proposed method got selected.

On an Euler-Lagrange equation for color to grayscale conversion

2019 ◽  
Vol 2019 (1) ◽  
pp. 95-98
Author(s):  
Hans Jakob Rivertz
Keyword(s):  
Variational Inequality ◽  
Color Image ◽  
Lagrange Equation ◽  
New Method ◽  
Grayscale Image ◽  
Structure Tensors

In this paper we give a new method to find a grayscale image from a color image. The idea is that the structure tensors of the grayscale image and the color image should be as equal as possible. This is measured by the energy of the tensor differences. We deduce an Euler-Lagrange equation and a second variational inequality. The second variational inequality is remarkably simple in its form. Our equation does not involve several steps, such as finding a gradient first and then integrating it. We show that if a color image is at least two times continuous differentiable, the resulting grayscale image is not necessarily two times continuous differentiable.

Deep Neural Network Classification of I-123 Ioflupane SPECT

2020 ◽  
Author(s):  
David T. Wang ◽  
Brady Williamson ◽  
Thomas Eluvathingal ◽  
Bruce Mahoney ◽  
Jennifer Scheler
Keyword(s):  
Neural Network ◽  
Deep Neural Network ◽  
Neural Network Classification
Sign in / Sign up

Export Citation Format

Share Document