Good morning. I'm writing a scientific paper and I need to compare the performance of several recognition algorithms/models (such as Conformer, which I read to be supported by openspeech) on the UA-Speech dataset (http://www.isle.illinois.edu/sst/data/UASpeech/).
UA-Speech is a dataset for users affected by disartria. I wonder whether it is possible to test the recognition performance of the
openspeech framework on the UA-Speech dataset. Assuming that such operation can be theoretically performed, what steps would be necessary to achieve such result ? How can I elaborate UA-Speech dataset using Openspeech source code base?
Good morning. I'm writing a scientific paper and I need to compare the performance of several recognition algorithms/models (such as Conformer, which I read to be supported by openspeech) on the UA-Speech dataset (http://www.isle.illinois.edu/sst/data/UASpeech/).
UA-Speech is a dataset for users affected by disartria. I wonder whether it is possible to test the recognition performance of the openspeech framework on the UA-Speech dataset. Assuming that such operation can be theoretically performed, what steps would be necessary to achieve such result ? How can I elaborate UA-Speech dataset using Openspeech source code base?
Thanks in advance for your help.