I'm working with generation of deepfakes for Bangla language. While inferencing, I'm providing source audio of a speaker. To generate the deepfakes of remaining speakers using this, the output is not proper. The words spoken by source speaker are only audible in all deepfakes of other speakers.
I'm working with generation of deepfakes for Bangla language. While inferencing, I'm providing source audio of a speaker. To generate the deepfakes of remaining speakers using this, the output is not proper. The words spoken by source speaker are only audible in all deepfakes of other speakers.