3 . EXPERIMENTAL RESULTS 3 . 1 . Development Test Set
ثبت نشده
چکیده
A new scoring algorithm has been developed for generating wordspotting hypotheses and their associated scores. This technique uses a large-vocabulary continuous speech recognition (LVCSR) system to generate the N-best answers along with their Viterbi alignments. The score for a putative hit is computed by summing the likelihoods for all hypotheses that contain the keyword normalized by dividing by the sum of all hypothesis likelihoods in the N-best list. Using a test set of conversational speech from Switchboard Credit Card conversations, we achieved an 81% figure of merit (FOM). Our word recognition error rate on this same test set is 54.7%.
منابع مشابه
Developments in continuous speech dictation using the ARPA WSJ task
In this paper we report on our recent development work in large vocabulary,American English continuous speech dictation. We have experimented with (1) alternative analyses for the acoustic front end, (2) the use of an enlarged vocabulary so as to reduce the number of errors due to out-of-vocabulary words, (3) extensions to the lexical representation, (4) the use of additional acoustic training ...
متن کاملFault Classes and Error Detection Capability of Speciication Based Testing
Some varieties of speciication based testing rely upon methods for generating test cases from predicates in a software speciication 1] 2] 3] 4] 5] 6] 7]. These methods derive various test conditions from logic expressions, with the aim of detecting diierent types of faults. Some authors have presented empirical results on the ability of speciication based test generation methods to detect failu...
متن کاملSimulation Calibration of an Extensive Air Shower array, by a logged data set of its small prototype
Alborz observatory is an EAS array in the heights of Alborez mountain chain near Tehran. For the development of the array, more number of detectors is inevitable. The managing the financial resources and achieving the highest efficiency of the array is important for the project. Therefore, Water Cherenkov Detectors has been used for 9 months in the same way as the previous experiment with Scint...
متن کاملThe comparison of the effectiveness of Fredrickson’s positive thinking and Dweck’s mind set program on psychological capital of ninth-grade gifted girl students with test anxiety of Isfahan
Introduction: The present study aimed to investigate the comparison of the effectiveness of Fredrickson’s positive thinking and Dweck’s mind set program on psychological capital of ninth-grade gifted girl students with test anxiety of Isfahan. Method: This study is quasi-experimental with pretest-posttest with control group. The population of the study consisted all 450 gifted female student...
متن کاملAutomated Maintainability of TTCN-3 Test Suites Based on Guideline Checking
Similar to software development, the test development must be accompanied with a set of rules specifying how to write tests. They are grouped together into a document called guideline. Guidelines are especially necessary for large test specifications involving many developers and have the goal to reduce the effort of the overall development. So far, no universal guidelines for the TTCN-3 langua...
متن کامل