Deep Neural Networks for End-to-End Optimized Speech Coding

Simple item page
dc.contributor.advisorJacobs, Daviden_US
dc.contributor.authorKankanahalli, Sriharien_US
dc.contributor.departmentComputer Scienceen_US
dc.contributor.publisherDigital Repository at the University of Marylanden_US
dc.contributor.publisherUniversity of Maryland (College Park, Md.)en_US
dc.date.accessioned2017-09-14T05:49:44Z
dc.date.available2017-09-14T05:49:44Z
dc.date.issued2017en_US
dc.description.abstractModern compression algorithms are the result of years of research; industry standards such as MP3, JPEG, and G.722.1 required complex hand-engineered compression pipelines, often with much manual tuning involved on the part of the engineers who created them. Recently, deep neural networks have shown a sophisticated ability to learn directly from data, achieving incredible success over traditional hand-engineered features in many areas. Our aim is to extend these "deep learning" methods into the domain of compression. We present a novel deep neural network model and train it to optimize all the steps of a wideband speech-coding pipeline (compression, quantization, entropy coding, and decompression) end-to-end directly from raw speech data, no manual feature engineering necessary. In testing, our learned speech coder performs on par with or better than current standards at a variety of bitrates (~9kbps up to ~24kbps). It also runs in realtime on an Intel i7-4790K CPU.en_US
dc.identifierhttps://doi.org/10.13016/M2M03XX8H
dc.identifier.urihttp://hdl.handle.net/1903/20035
dc.language.isoenen_US
dc.subject.pqcontrolledArtificial intelligenceen_US
dc.subject.pqcontrolledAcousticsen_US
dc.titleDeep Neural Networks for End-to-End Optimized Speech Codingen_US
dc.typeThesisen_US

Files

Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
Kankanahalli_umd_0117N_18451.pdf
Size:
2.27 MB
Format:
Adobe Portable Document Format
Download

Collections

UMD Theses and Dissertations
Computer Science Theses and Dissertations