Effective speech not only comes from confidence but also skills. For this project, we decided to focus on coming up with an effective solution for people who are suffering from presentations or speeches. Since our team members are students from various countries, we agreed on the fact that finding effective guidelines or helpers is difficult while preparing for an urgent speech or presentation. We realized that the majority of the people are struggling with seeking immediate help or guidelines at home or other places. With that, we thought that an easily accessible platform is indispensable for delivering direct support. By “Speechly: Speech + Skillfully,” we aim to provide a sufficient environment for the users as a digital speech helper.
What it does
Speechly’s primary function is the function that allows the users to receive direct feedback and to practice their own speech. As the user speaks in front of the camera, it detects the user’s posture and speech. The text of the script the user typed in and what the user is saying appear on the screen. Speechly composes a machine learning model trained on our body gestures, such as eye contact and shoulder, as well as a speech-to-text system to foster the user’s demonstration skills. In addition, we have extra guideline videos featured on our website.
How we built it on
In prior to the actual coding process, we researched resources such as tip videos that Speechly can have to effectively help the intended users. We mainly used Repl.it and Figma to collaborate virtually for the leading web features. Also, we used Firebase to facilitate the process of signing up and collecting diary scripts into the database.
Challenges we ran into
Some of the biggest challenges that affected our team were the difference in time zones and collaborating through a virtual environment. Since we had to overcome the time zone difference for both the collaboration and the workshops, which was about 15~17 hours max among team members, we had to frequently stay up until midnight in order to communicate and keep track of the project. In the end, we were able to showcase the maximum potential of Speechly despite the harsh environment and the short time limit.
What is next for “Speechly”
In the future, we aim to add additional functions such as a grammar checking function for the users. Also, we would like to expand Speechly to a chrome extension that will enable the possibility of the users more effectively approaching it. We are hoping that Speechly will be able to support users around the globe and positively influence them.