Datasets:
metadata
license: apache-2.0
task_categories:
- automatic-speech-recognition
language:
- ki
tags:
- agriculture
- kenya
size_categories:
- 10K<n<100K
This dataset is obtained as part of AIEP prject by Digital Green and Karya from the extension workers, lead farmers and farmers.
Process of collection of data: Selected users were given the option of doing a task and getting paid for it. The users were supposed to record the sentence as it appeared on the screen. The audio file thus obtained was validated matched with the sentences to fine tune the model.
Also available are the python script that helps in processing and splitting the data into train and test sets.