Project Info

ASL Live Translator

Devpost

Inspiration

The inspiration behind the ASL Live Translator came from a desire to bridge communication gaps between the hearing and deaf communities. Many people struggle with language barriers, especially when it comes to sign language. This website aims to make communication more accessible and inclusive.

What it does

The ASL Live Translator uses a webcam to capture American Sign Language (ASL) signs, which are then converted into live captions. This allows users to communicate seamlessly with individuals who are deaf or hard of hearing by translating ASL into text.

How we built it

We built the website using Next.js, incorporating: TensorFlow.js and the Handpose model for recognizing ASL signs. Fingerpose for gesture estimation. Webcam feed integration using the react-webcam library. Text-to-Speech integration for real-time speech conversion. The project involves custom components like WebcamFeed, LiveCaptions, and SpeechSynthesis.

Challenges we ran into

Webcam integration was a bit tricky, especially with handling different camera states and permissions. ASL recognition accuracy was challenging, as the model had to accurately capture and identify signs in varying lighting conditions and different sign variations. Ensuring real-time performance was a struggle with the amount of data processed during sign detection and translation.

Accomplishments we're proud of

Successfully integrating real-time ASL translation to captions using machine learning. Making the website responsive, clean, and accessible for different devices. Ensuring the translation happens almost immediately with a user-friendly interface.

What we learned

Gained hands-on experience with TensorFlow.js and machine learning models for real-time applications. Learned more about gesture recognition and its limitations in real-world usage. Improved problem-solving skills related to real-time video processing and webcam integration.

What's next

Expanding the range of ASL signs supported, including more complex gestures and phrases. Integrating voice recognition to allow for a two-way communication flow between the user and the application. Adding a feedback system to improve the accuracy of sign recognition based on user input and corrections.

Analysis

Compare with all teams

View

Metric

Figures cover GitHub contributors during the hackathon window. A co-authored commit counts in full for each author, so per-member totals add up to more than the whole-team figures.

Technology

Found in codeClaimed only
  • CSSIn code
  • Next.jsIn code
  • ReactIn code
  • Tailwind CSSIn code
  • TypeScriptIn code
  • JavaScriptClaimed

5 of 6 appear in the indexed code. 1 claimed on Devpost could not be matched to code, which may simply mean the tool leaves no trace in the repository.

AI coding agents

No AI coding agent signals were found in this repository.

Detected from committed agent config files and commit authorship. Absence of a signal is not proof an agent was unused.

Codebase size

Source size

14 KB

Source files

15

Counts recognized source files only; vendored directories, binaries and lockfiles are excluded, so this is smaller than the repository on disk.

0 stars