In the speech recognition industry, there is an everlasting need for evaluation of products in environments imitating real use cases. A wide-spread solution is to build a setup compliant with ETSI EG 202 396-1 standard, which defines a unified artificial laboratory environment to simulate real use scenarios of products to be tested. For space and cost reduction, a method is being developed to miniaturize the standard setup and simulate its behavior in a soundproof enclosure. In order to achieve high fidelity a number of spectral and temporal qualities of sound are measured in a laboratory and replicated in a box. The performance is evaluated using metrics specific to speech recognition.
(See document for exact affiliation information.)
https://aes2.org/publications/elibrary-page/?id=19528
(1153KB)
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 Join the AES. If you need to check your member status, login to the Member Portal.
