chunked/nltk_chunking/model_doc_clap/chunk_7.txt

In this paper, we propose a pipeline of contrastive language-audio pretraining to develop an audio representation by combining audio data with natural language descriptions.