Concealed Captioning

Landing Page
Video editing page to generate face blur and subtitles
Edited sample video page
Social Media post template

Inspiration

The idea for this project came to life after witnessing the protests in Iran. A large number of videos surrounding the protest surfaced online int he span of a few days to raise awareness about what was happening inside the country. Unfortunately, many people uploaded these videos without blurring the faces of fellow protesters, either because they did not know how to do it or did not have enough time. This led to many arrests, as there was clear video evidence of individuals protesting. We wanted to create something that would help people who lack video editing skills or time to still share content safely, without putting anyone’s life in danger, and help spread their messages across the world.

What it does

Our application allows users to upload multiple videos and edit them by adding subtitles and blurring faces. We also made it easy for users to post their edited videos directly to the social media platforms of their choice, saving them time and effort.

How we built it

We built the project by assigning each team member a specific role. One member was responsible for the frontend and UI, one worked on face detection and the last on subtitle generation. We collaborated closely to ensure smooth communication between the frontend and backend.

Challenges we ran into

One of our biggest challenges was adding subtitles to the videos. We had to directly embed them into the video pixels and ensure they were perfectly synchronized. This required extracting accurate timestamps and matching them correctly with the video frames.

Accomplishments that we're proud of

We are proud that we successfully implemented subtitles that stay in sync with the videos and support the user’s chosen language. This was also our first time working with YOLO-based face detection, so getting face blurring to work correctly was a major accomplishment.

What we learned

We learned how to manipulate video data and how to properly feed information into an AI model (YOLOv8) to achieve our goal of blurring faces. We also gained valuable experience in system design by structuring the data flow so that the frontend and backend could communicate efficiently and reliably.

What's next for Concealed Captioning

Adding user authentication and persistent backend storage so users can access their videos after leaving the website
Adding more AI-powered editing features, allowing users to describe what they want done to a video and having the system perform the edits automatically