I recently learned about the TSP speech database and was wondering if you ever heard of it? None of the papers I've read had mentioned it, so I wasn't sure if it was credible or just unknown.
I ask because I saw that someone had benchmarked Google's WebSpeech API with it and was curious if anyone else had already done so with Sphinx.
Thoughts on the matter?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Last years the amount of speech data available increased significantly and there are many databases you could test on.
The problem with the test on just a database is that it doesn't demonstrate the ability of the system to solve specific issues with speech recognition like robustness to accent. It makes more sense to setup more controlled experiments which demonstrate how you solve a problem than to test on some data collected somewhere.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi,
I recently learned about the TSP speech database and was wondering if you ever heard of it? None of the papers I've read had mentioned it, so I wasn't sure if it was credible or just unknown.
I ask because I saw that someone had benchmarked Google's WebSpeech API with it and was curious if anyone else had already done so with Sphinx.
Thoughts on the matter?
Last years the amount of speech data available increased significantly and there are many databases you could test on.
The problem with the test on just a database is that it doesn't demonstrate the ability of the system to solve specific issues with speech recognition like robustness to accent. It makes more sense to setup more controlled experiments which demonstrate how you solve a problem than to test on some data collected somewhere.