id	author	title	date	pages	extension	mime	words	sentence	flesch	summary	cache	txt
ap-1105	Rajnoha, J.	Multi-Condition Training for Unknown Environment Adaptation in Robust ASR Under Real Conditions	2009	5	.pdf	application/pdf	4428	238	51	The decrease by 29 % and 14 % in word error rate in comparison with clean speech training data was achieved for the non-adapted and adapted system, respectively. 20 50 100 200 500 1000 all time 65 s 2.8 min 5.7 min 10.9 min 26.9 min 57.2 min 7.8 h Table 3: Average amount of speech data for limited adaptation subsets ALL OFFICE CAR CLEAN NOISY HISNR LOSNR AVG avg avg CS0 CS1 CS0 CS1 CS0 CS1 CS0 CS1 CS0 CS1 CS0 CS1 CS0 CS1 CS0 CS1 noSS 8.32 14.72 3.47 8.14 7.03 35.17 4.0 11.07 13.09 18.79 7.03 10.2 11.88 17.65 12.18 7.83 16.53 SS 8.46 13.53 4.17 8.01 5.2 29.97 4.45 9.82 11.31 16.56 6.83 9.41 11.76 16.31 11.13 7.45 14.8 Imp. �1.68 8.08 �20.17 1.6 26.03 14.79 �11.25 11.29 13.6 11.87 2.84 7.75 1.01 7.59 8.66 4.82 10.48 Table 4: WER for different environmental conditions w/o and with ESS in front-end processing.	cache/ap-1105.pdf	txt/ap-1105.txt