Request for more information
#1
by
Cablecutter
- opened
Hi, could you provide some additional information about this dataset? Where was it derrived and how have you used it to train your models? What sort of success have you seen over baseline? I am working on a project to transcribe swedish interviews and am looking for guidance.
Hi, apologies for the delay, just seeing this right now. The project was for an assignment in our studies, and it was mostly experimental. The results were not the best, and produced a lot of mistakes. Could only transcribe very simple sentences.
Also the "augmented" was an experiment in applying a filter to the spectogram image, that was not successful.