Technical report

MATĚJKA Pavel, GLEMBEK Ondřej, PLCHOT Oldřich, SCHWARZ Milan, CIPR Tomáš, CUMANI Sandro, KUDLA Radim, SZŐKE Igor, SVOBODOVÁ Marie, MALÝ Květoslav and ČERNOCKÝ Jan. BUT HASR'12 Experience: Are developers of SRE Systems naive listeners?. Brno: Faculty of Information Technology BUT, 2013.
Publication language:english
Original title:BUT HASR'12 Experience: Are developers of SRE Systems naive listeners?
Title (cs):HASR'12 zkušenost na VUT: Jsou vývojáři SRE systémů naivní posluchači ?
Place:Brno, CZ
Publisher:Faculty of Information Technology BUT
automatic speaker recognition, human assisted speaker recognition, HASR, i-vector
NIST 2012 Human Assisted Speaker Recognition (HASR) task are described in this paper. The scores of an automatic i-vector based system and 10 na¨ıve listeners were fused. The results show that an automatic system performed generally better than human listeners, with the exception of an engineer that has never done such analysis before. The conclusions of this study should be taken with care due to low number of trials.
   author = {Pavel Mat{\v{e}}jka and Ond{\v{r}}ej Glembek and
	Old{\v{r}}ich Plchot and Milan Schwarz and Tom{\'{a}}{\v{s}}
	Cipr and Sandro Cumani and Radim Kudla and Igor Sz{\H{o}}ke
	and Marie Svobodov{\'{a}} and Kv{\v{e}}toslav Mal{\'{y}} and
	Jan {\v{C}}ernock{\'{y}}},
   title = {BUT HASR'12 Experience: Are developers of SRE Systems naive
   pages = {1--4},
   year = {2013},
   location = {Brno, CZ},
   publisher = {Faculty of Information Technology BUT},
   language = {english},
   url = {}

Your IPv4 address:
Switch to IPv6 connection

DNSSEC [dnssec]