Extract unique speaker voice sample from converstional audio data using deep learning.
Much information can be extracted from recorded voice data such as the speaker's identity, the conversation between speakers, and the exact time stamp of when the speaker speaks. Manually, lengthy recorded voice data information extraction process takes significant time and energy.
Speaker diarization overcame this problem by dividing pieces of recorded voice data into groups based on the speaker's identity while specifying when the speaker speaks. A speaker diarization system was made by performing segmentation on voice data, extracting unique voice features, and cluster the extracted voice features.
Jokowi
Speaker 1 sample voice
Prabowo
Speaker 2 sample voice
MC
Speaker 3 sample voice
© Azka Radinka. 2023