ud-nlp/

call-center-audio

Dataset is a large audio dataset containing 13,000+ hours of real-world customer service calls from global call centers, featuring 90%+ unique speakers and time-stamped transcripts for accurate speech recognition, speaker diarization, and conversational AI model training. - Get the data

	Dataset characteristics:




	

Characteristic Data

Description Audio of real customer service calls

Data types Audio

Tasks Speech… See the full description on the dataset page: https://huggingface.co/datasets/ud-nlp/call-center-audio.