id	author	title	date	pages	extension	mime	words	sentence	flesch	summary	cache	txt
fcis-22002	Li , Yian; Huang, Yibo	End-to-End Speech Hash Retrieval Algorithm based on Speech Content and Pre-training	2024	7	.pdf	application/pdf	6497	350	53	In order to be able to perform speech retrieval in noisy environments, we manually added Gaussian white noise to the test dataset and constructed speech retrieval tasks with different signal-to-noise ratios (SNRs), including 30 db, 25 db, 20 db, 15 db, and 10 db. Although the use of transcription technology has largely improved the accuracy of speech retrieval, its efficiency and accuracy are highly dependent on advanced speech recognition technology.	cache/fcis-22002.pdf	txt/fcis-22002.txt