After receiving a Quinnipiac University grant in June, Chetan Jaiswal, associate professor of computer science, is creating technology capable of instantly translating two voices on a phone call in as many as 100 languages.
Jaiswal is working to create the app with Peter Zegarek, a sophomore 3+1 computer science major, and they plan to complete the first prototype by the end of the 2023-24 academic year. Users will eventually be able to apply the program to Zoom meetings and YouTube videos.
The translator does not currently know any languages because it is in the developmental stage. Jaiswal said he has been studying language learning models that already exist but has not yet started the data collection process. Members of the Quinnipiac community, such as faculty and students will create the data collection research team.
Artificial intelligence, which the project is created from, uses computers and machines to mimic problem-solving and decision-making skills.
Jaiswal teaches about AI in his database and programming classes to Quinnipiac students. He said he is passionate about AI technology because it can make daily actions easier for him through computer programs.
“I introduced different concepts here and there so students get excited about it and try to learn more about it,” Jaiswal said.
Jaiswal is among 10 Quinnipiac University faculty members who received grants from the Faculty Scholarship and Creative Works Impact Fund administered by the Office of the Provost. Quinnipiac administrators implemented the fund to advance opportunities for full-time faculty as scholars, practitioners and creators.
The donor-funded grants are worth between $4,000-$5,000. Jaiswal said the money from the grant will fund student stipends and materials.
Zegarek joined the project after Jaiswal brought up a research opportunity during his class. He talked to him after class and the two decided to work together. He said he has been wanting to get more into the computer science field.
“So, because I felt that I had a good foundation to go off of, I wanted to start to expand out and actually work on something that can impact and help people,” Zegarek said.
Jaiswal said using human translators and platforms such as Google Translate requires more time than AI because they do not allow people to communicate without delays.
“Now this doesn’t only apply to students in Zoom lectures or in-person lectures, it also applies to anywhere where communication is important,” Jaiswal said.
Anyone will be able to use the translator once it’s available, Jaiswal explained, whether it is in a job setting or talking to a friend on the phone.
“The project goes way beyond the scope of the grant itself,” Jaiswal said.
Jaiswal plans to teach the AI to translate American Sign Language by having one person use ASL through the camera on their phone. Then, the AI will play the translation back to another person on the other line. He said Quinnipiac students will be able to utilize the translator once it is completed to listen to class lectures or with their patients if they are in the health sciences.
“I think it can help us go beyond the normal student body,” Jaiswal said. “We can have students that are not native English speakers and still can be part of (Quinnipiac) because they can get the lecture in the language of their choice,” Jaiswal said.
Zegarek emphasized that AI translators will allow people to speak more easily with others that speak different languages, such as family members who live in other countries.
“I think this project is very important because it allows us to help bridge gaps between people who may have a lack in communication between each other,” Zegarek said.
