Project Wolf

Great opportunity for the Low vision and Blind people to work on the project _ WOLF . Project will include Audio and Transcription both. Deliverables We will provide Company with: - Data necessary to run the tests for each of the languages: o European languages: English Great Britain, FR-French IT-Italy, DE- Danish ES-Spanish Potentially: English -US - Per language, the data comprises: - Audio samples: ▪ Each audio sample is a sentence, with the exception of one which is a short paragraph. ▪ Number of sentences: • Min: 256 sentences (2 sets of 128 sentences). • Max: 484 sentences (2 sets of 242 sentences). ▪ Text of each sentence will be provided as a ground truth to compute WERs. ▪ The sentences may be synthesised at fast speech rates (from 170WPM to 750WPM). ▪ Average length of audio samples is less than 2 seconds. o One question about how easy it was to understand the text, per sentence. o Instructions per test screen. o Set of questions to survey participants at the beginning of the test: ▪ Age: ▪ Gender: ▪ Vision status: • Blind. • Low vision. ▪ At what age did you become blind or low vision?: • 0 (from birth) • [age # to be filled by participant] ▪ How do you normally access audio content on your computer? • headphones. • laptop speakers. • external speakers. • other. - How often do you listen to computer-generated speech? • Less than 1 hour per week. • Between 1 and 10 hours per week. • Between 10 and 30 hours per week. • More than 30 hours per week. - How often do you use screen reader technology? • Less than 1 hour per week. • Between 1 and 10 hours per week. • Between 10 and 30 hours per week. • More than 30 hours per week. - Which kind of computer-generated voice do you usually prefer? • female. • male. • no preference. - Set of questions to survey participants at the end of the test: ▪ How easy or difficult was this evaluation to complete? • Very Difficult • Somewhat Difficult • Neither difficult nor easy • Somewhat Easy • Very Easy ▪ Would you like to add any suggestions or comments about the evaluation? - All Participants must be native speakers of the language of the voice being tested. - All participants must be visually impaired (either with low vision or blind). - Participants should not need any assistance to type (they should by able to type on a keyboard). - Participants should be screen reader users.