Automated recovery of damaged audio files using deep neural networks

Heo, Hee-Soo; So, Byung-Min; Yang, Il-Hyung; Yoon, Sung-Hyun; Yu, Ha-Jin

doi:10.1016/j.diin.2019.07.007

Cited by 8 publications

(6 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…В исследовании [2] предложены методы восстановления аудиофайлов, основанные на нейронных сетях с прямой связью и долговременной памятью, а также описаны возможности автоматизации задачи восстановления сигнала с помощью предлагаемых методов восстановления, использующих глубокие нейронные сети.…”

Section: анализ состояния предметной областиunclassified

Method for Restoring Audio Signal Using Convolutional Neural Networks

Morozova

Rakitskiy

2023

jour

View full text Add to dashboard Cite

This paper describes a research of the restoring distorted audio signal possibility. Based on the previously obtained results of using deep machine learning methods, the concept of a neural network to correct a distorted audio signal has been developed. On the basis of the originally obtained results, several new neural network architectures were developed, focused on the audio signal restoring. The paper contains descriptions of the developed architectures with a theoretical substantiation of the possibility of their application. The presented architectures were tested to solve the problem of restoring the part of a specifc instrument in a musical composition where it was removed. The results of testing the developed architectures of neural networks are presented in several forms.

show abstract

Section: анализ состояния предметной областиunclassified

Method for Restoring Audio Signal Using Convolutional Neural Networks

Morozova

Rakitskiy

2023

jour

View full text Add to dashboard Cite

show abstract

“…With the development of machine learning techniques, deep learning-based methods are also being applied to file recovery (Mohammad & Alqahtani, 2019). Heo et al (2019) described a long short term memory-based approach for audio file recovery. In contrast to traditional methods, In contrast to traditional file recovery methods, Heo et al's method aims to recover audio that can be distinguished by the human ear, rather than recovering the original exact data as in traditional methods.…”

Section: Research On File Recoverymentioning

confidence: 99%

“…Different from MFR, CFR does not rely on metadata. It leverages syntactic signatures (e.g., file header-footer pairs) (Tang et al, 2016), semantic structures (e.g., explicit control flow paths within a binary executable) (Hand et al, 2012), heuristic technologies (Garfinkel & McCarrin, 2015;Gladyshev & James, 2017;Pal et al, 2008), timestamps (Nordvik et al, 2020;Portera et al, 2021) or deep learning technologies (Heo et al, 2019;Mohammad & Alqahtani, 2019) to restore files. Unlike MFR, which can precisely recover a file under the "direct guidance" of metadata, CFR "indirectly infers" which data blocks belong to the file to be recovered.…”

Section: Introductionmentioning

confidence: 99%

Golden Eye

Zhang

Chen

Zhu

2022

International Journal of Digital Crime and Forensics

View full text Add to dashboard Cite

File systems are important sources of intelligence information and digital evidence. They have long attracted the interest of researchers in recovering files that are deleted from a hard disk. Existing file recovery studies rely heavily on an operating system (OS). However, it is often encountered that OS services are not available, making existing file recovery approaches unusable. To address this issue, the authors design and implement an OS-independent file recovery algorithm named Golden Eye (GE) by targeting the EXT4 file system. Fed the raw image obtained from a (sanitized) hard disk, GE can automatically recover any designated file or even the whole EXT4 file system. GE is based on the understanding of the file disk layout of EXT4 and does not need any support from additional hardware or software. Experimental results prove the feasibility and correctness of GE. This work not only solves the OS dependency problem that most existing file recovery work suffers from but also reveals the fact that even sanitized hard disks are still at risk of leaking sensitive data.

show abstract

“…A DNNs classifier can directly input all information into neural networks and automatically identify the implicit connections within the data. With a deep and wide neural network structure, the DNNs classifier has substantial information analysis abilities [43]. The process of extracting feature vectors using the TFE algorithm in Section 2.3 can be essentially understood as a process of mapping the raw data to highdimensional vector space.…”

Section: The Twin-window Feature Extraction (Tfe) Processmentioning

confidence: 99%

Non-intrusive classification of gas-liquid flow regimes in an S-shaped pipeline riser using a Doppler ultrasonic sensor and deep neural networks

Nnabuife

Kuang

Whidborne

et al. 2021

Chemical Engineering Journal

View full text Add to dashboard Cite

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

show abstract

Automated recovery of damaged audio files using deep neural networks

Cited by 8 publications

References 3 publications

Method for Restoring Audio Signal Using Convolutional Neural Networks

Method for Restoring Audio Signal Using Convolutional Neural Networks

Golden Eye

Non-intrusive classification of gas-liquid flow regimes in an S-shaped pipeline riser using a Doppler ultrasonic sensor and deep neural networks

Contact Info

Product

Resources

About