Spent our childhood in Korea and Taiwan, Jeana and Angel both attended international schools and are now able to speak English fluently. However, others (from developing countries' local school students to visually impaired teenagers) are not as fortunate as Angel and Jeana. The underprivileged communities never had the chance to develop their English or second language skills to open up more future opportunities, making it harder for developing countries to continue to grow as a nation. This tool was built with the hope to make it easier for everyone to communicate with other people around the world and learn another language. With this, we also help nurture a more sustainable community.
What it does
SPEC-tacular Translator uses APIs and Cloud Services in order to perform realtime translation during a conversation, easing tourists' language barriers and supporting students' ability to learn a second language. More importantly, we added a braille translation to extend the benefits to visually impaired patients, hoping to ameliorate their living standards and community involvement.
The SPEC-tacular Translator supports realtime conversation between multi-parties, where the user selects the incoming and outgoing language and specifies the recipient. The conversation continues until a trigger word is heard, such as "exit".
How we built it
In order to perform realtime analysis, we utilized Google's Cloud APIs in addition to Firebase. Additionally, we personally implemented braille translation through dictionary mapping. We then used Flask to improve the user experience.
Challenges we ran into
One challenge we ran into was syncing multiple translations occurring at the same time. We resolved this issue by change our method of storing from FireBase to FireStore.
Accomplishments that we're proud of
Despite being a two-person team and new to the software/tools, we managed to learn about FireBase and Google Cloud in a short period of time and create a working project.
What's next for SPEC-tacular Translator
We would like to incorporate hardware for the speech-to-braille translation. Currently we did not have the specific hardware we wanted to use so the braille translation is displayed on the screen temporarily.