Recent research projects have attempted to realize high realism of reproduced sound fields. We propose an innovative method, an automatic scene discrimination, for a user-centric rendering of enhanced realism in the home theater application. The method categorizes various scenes of movie contents based on their audio characteristics, and applies a pre-determined signal processing for each category in real-time. The training model was able to classified new movie contents with a 71% correct ratio. Listeners reported that this scene-based adaptive sound processing bring higher realism.

Authors: Yuyama, Yuta; Kim, Sungyoung; Okumura, Hiraku
Affiliations: Yamaha Corporation, Shizuoka, Japan; Rochester Institute of Technology, Rochester, NY, USA(See document for exact affiliation information.)
AES Conference: 2018 AES International Conference on Spatial Reproduction - Aesthetics and Science (July 2018)
Paper Number: EB1-6
Publication Date: July 30, 2018 Import into BibTeX
Session Subject: Sound Field Reproduction; Multichannel Audio; Home Theater System; Scene Classification; Machine-Learning; SVM; Feature-Extraction
Permalink: https://www.aes.org/e-lib/browse.cfm?elib=19646

Click to purchase paper as a non-member or login as an AES member. If your company or school subscribes to the E-Library then switch to the institutional version. If you are not an AES member and would like to subscribe to the E-Library then Join the AES!

This paper costs $33 for non-members and is free for AES members and E-Library subscribers.

Learn more about the AES E-Library

E-Library Location:   /conf/2018/spatial/spatial_reproduction_2018_paper_63.pdf

Start a discussion about this paper!