Wipro Technologies' training managers sought a fast and cost-efficient solution for adding and updating voice narration audio in their Web-based training courses. Tuval Software Industries' Speech-Over text-to-speech voice narration software fulfilled their requirements by replacing voice-over recording with automatic generation of narration from text, making it fast and easy for presentation authors to add and update narration. Costs in content creation and administration were reduced as Wipro successfully rolled out an initial set of six courses with embedded voice generated from Speech-Over.
As semiconductor technology and processes rapidly change, there is a continuous need to enhance the competency of the team in the VLSI/System Design area. Traditionally, training courses were administered by experts in a classroom setting, but this model did not scale up with the rapid growth in team size, and was further exacerbated by non-availability of expert's time on a periodic basis. Wipro's development centers are spread across the world, making it difficult to conduct classroom training simultaneously at every location. Hence, Wipro training executives started looking at an e-learning model which addresses all these needs without compromising the quality much: Experts would record their voices along with the presentation, enabling trainees to go through the lessons offline and discuss any questions with the expert in a discussion forum or chat room.
"We wanted to make it easier to update and revise our courses by not having to hunt for the original professional voice whenever changes are to be done and, even better, not to have to use a professional voice at all to save cost," explains Raj Kothandaraman, practice head of Wipro Technologies' VLSI Design & Verification Practice Group. "We were also looking at Text-To-Speech (TTS) conversion tools to embed audio into Microsoft PowerPoint for ease of content creation. Tuval Software Industries’ Speech-Over text-to-speech software solution met these requirements very well." Using Speech-Over software in PowerPoint design mode, course developers use a simple form-based user interface to quickly and easily accept narration script text and automatically generate professional text-to-speech voices that speak the text. All SAPI-compatible vendor voices can be used, but Wipro chose AT&T Natural Voices. Using Speech-Over, developers also can record their own voices and import pre-recorded sound files. With its new Voice Effect Marshalling technology, Speech-Over lets course developers automatically synchronize slide narration with screen animations, allowing them to describe and explain text and graphic objects on a slide in any order and create a highly effective multi-media training movie.
After six months of operation with Speech-Over, 15 course developers currently work with the product, and the company plans to expand to 30 users or more. It took Wipro developers just a few days to get familiarized and start working. Speech-Over's help file and the tutorial demos made this task even simpler. As the work progressed, there were queries and issues raised, and this was satisfactorily resolved through e-mail by Tuval Software Industries support staff in a timely fashion. In addition, Wipro's staff made useful suggestions for new Speech-Over features, which were implemented in subsequent versions. Speech-Over helped significantly reduce efforts in learning content creation and administration. Wipro has rolled out an initial set of six courses with embedded voice generated from Speech-Over to a positive user response. Wipro course developers are now working to produce more natural and conversational audio for an even better learning experience.
Joel Harband, a software developer, is the founder and CEO of Tuval Software Industries, Ltd. For more information, visit www.speechover.com. To learn more about its customer, Wipro Technologies, visit www.wipro.com.