Chulalongkorn University Unveils Gowajee: A Revolutionary Thai Speech-Recognition AI Delivering Native-Like Accuracy

Published September 22, 2023

The demand for AI-based speech recognition software has grown exponentially over the past few years. However,according to 66% of respondents in a study conducted in 2020, with over 7,000 languages spoken worldwide having uncountable dialects and accents, voice tech adoption can be problematic. This is especially true for Thai speakers who face challenges in using foreign-made AI that often fail to comprehend the nuances of the Thai tone of voice. The inability of standard AI systems to accurately interpret the complexities of the Thai language structure has driven the need for a genuine Thai speech recognition solution.

Dr. Ekapol Chuangsuwanich and his team at the Department of Computer Engineering, Faculty of Engineering, Chulalongkorn University, recognized this pressing need and embarked on the ambitious project of developing Gowajee. The team successfully compiled an impressive five-thousand-hour sound database by creating the most extensive Thai language database through years of rigorous data collection, including conversational logs, emotional speaking, and text readings. This vast compilation forms the foundation of Gowajee's exceptional accuracy in transcribing Thai speech.

Gowajee excels in three main features, including Automated Speech Recognition (ASR), Text-to-Speech (TTS), and Automatic Speaker Verification (ASV). ASR enables seamless speech conversion into text, making transcription tasks significantly more efficient and accurate. TTS, akin to widely-known virtual assistants, produces natural speech with a genuine Thai tone, thanks to the more ext

ensive Thai database. ASV offers secure identity verification through voice, which is ideal for call centers and time-stamped interactions.

Their AI has already found successful applications in various domains, particularly in call centers, where its error rate stands at an impressive 9%, compared to the 15% seen with other AI systems. Beyond call centers, Gowajee has a

lso shown its potential to contribute to essential areas such as DMIND, which aids in screening patients with depression by interpreting voices and decoding emotions.

Gowajee's versatility extends further as it holds promise for diverse applications, such as acting as a dental assistant, detecting stroke risks through slurred speech, acting as a life coach, assisting the hard-of-hearing, and much more.

One of the most distinguishing features of Gowajee is its unparalleled commitment to data safety. Unlike many transcription and speech recognition programs that store data on external servers or users' computers, Gowajee ensures data security by storing all information on the user's database. This critical feature makes it ideal for organizations prioritizing data confidentiality, including banks.

"AIs aren't that disrupting to our lives. We are disrupting ourselves. Aging societies, a shortage of working-age labor are making it necessary for us to create technologies to substitute what we can't find humans to do." Dr. Ekapol also concluded by saying, "I'm not expecting that my work is going to be helpful to the aged of today, but I'm thinking that in the future, when I reach an old age, I will be making use of these technologies."

As societies transition into aging populations, implementing speech recognition technologies will improve the overall quality of life and provide invaluable assistance in various aspects of daily living.

To experience the groundbreaking Gowajee Thai speech recognition AI and witness its transformative capabilities, visit the official website at

Chulalongkorn University
Phayathai Road, Pathumwan, Bangkok 10330 Thailand
Tel: +66 2218 3280

Release ID: 747971

Prodigy Press Wire

Your brand has a story. Prodigy Press Wire will help bring it to life.