Azure Warrenfeltz is fluent in Japanese and Spanish. She also can understand bits of French, German, Arabic and Italian, and she soon hopes to learn some Mandarin Chinese.
Azure is 4 years old.

In today's globalized world, Azure is one of many young American children whose parents insist her education include foreign languages.

"It's such a global environment now, you never know what you might need," says Azure's mother, Julie Warrenfeltz, "I wanted to make sure she had every tool and every benefit at her disposal.
Julie Warrenfeltz’s position is not a surprising one. The global market for 2nd language acquisition is US$12 billion and e-learning is an estimated US$1 billion as of 2007. Due to globalisation and the increase in international trade, tourism and immigration, these figures will undoubtedly increase. As such, the acquisition of multilingual skills has never been more important. With the increase in a demand for multilingualism, comes the supply of convenient and easily available language learning services.

The Solution

In order to provide the people with an easy and convenient language learning service, researchers have been working hard on creating the ABACUS. This is a state-of-the-art package speech training program that is a combination of award-winning developments speech and dialogue processing researches. The ABACUS aims to at building a multilingual, speaker independent phonetic speech recognition engine which is capable of recognising continuous and naturally spoken sentences or phrases.

The ABACUS program is a all-encompassing, ground-breaking invention which is not only multilingual and speaker independent. It also uses the novel “bag-of-sounds” technique and has the capacity to support a very large vocabulary. This means that the program can perform direct matching between any utterances from the learners and their model sound patterns. This would enable the efficient analysis of pitch, intonations and accents. If desired, a comparison score can also be generated based on phonetic and prosodic models. The most attractive of all is its easy availability for application developments, which is made able by a set of comprehensive APIs.

Easy Application

• Complementary Service to Language Schools

The ABACUS speech engine is a medium that is far more interactive than the traditional mediums available in the market today. Unlike published books and audio material, the ABACUS provides real-time feedback provided by speck comparisons that can be incorporated with rich 3D animations and programmed logic to match the user’s progress.

• Mobile Learning – Language Learning On the Go

Due to the major developments in portable technologies, which have led to the influx of mobile devices such as Smartphones, Palmtops, Tablet PCs and Pocket Pcs, opportunities for learning on the go have arisen. Students may now access advanced technological teaching services for learning as and when they want to, without worries of location.

• Language Learning Softwares

As the ABACUS Speech Engine is consolidated into an API format, it can easily be incorporated into language learning softwares. The ABACUS will provide a whole new dimension into learning and will thus set the language learning software apart from its competitors. In addition, a high-tech speech training suite will provide the students a new language learning experience.

ABACUS is also an award-winning speech engine. It won the 1st place in 17 out of 36 categories in the National Institute of Standards and Technology (NIST, USA) 2007 Language Recognition Evaluation and 2nd place in the NIST 2007 Rich Transcription International Benchmarking.

For enquiries or explore collaboration, please contact:
Industry Development Department
Tel: 65 6874 8399
Fax: 65 6775 9923
Email: inddev@i2r.a-star.edu.sg