Real-time fingerspelling video recognition achieving 92.7% letter accuracy on FSBoard
✌ Annotated dataset of Auslan Alphabet Images
🧏 Detection of the alphabets in American Sign Language (ASL).
Engineering thesis: a real-time ASL fingerspelling recognition desktop app using Python, PySide6, MediaPipe and OpenCV, with text and speech output
Bachelor’s Degree Final Project - Research on Sign Language Translation, advised by Prof. Yoav Goldberg and Mr. Amit Moryossef
Detect and Translate American Sign Language (ASL) fingerspelling into text.
Computer Vision powered sign language fingerspelling translator using MediaPipe hand tracking and a landmark classifier, spoken aloud through Web Speech API.
A collaborative AUT project by a 3-member team delivering a privacy-preserving NZSL/BSL fingerspelling recognition system using MediaPipe skeleton keypoints and routed machine learning branches with a real-time FastAPI dashboard.
Real-time NGT (Dutch Sign Language) fingerspelling learning app powered by a ResidualMLP and Bidirectional LSTM, with MediaPipe hand tracking and a FastAPI/WebSocket backend.
Real-time ASL fingerspelling: windowed MLP training, int8 TFLite deployment, and Unity demo.
A Web Based System that is made for a sign language translator model that is created using GLCM-CNN. This Site also features a fingerspelling page that teaches about signing the alphabet
Offline ASL fingerspelling trainer: reference, flashcards and quizzes. Next.js, runs fully in the browser, no account.
Reproducible part-aware temporal keypoint models and baselines for cross-signer Russian fingerspelling on Bukva.
Deep-learning ASL fingerspelling-to-text recognition using MediaPipe hand landmarks, Temporal CNN, BiGRU and CTC.
A system that recognizes and provides feedback on the hand movements and position involved in fingerspelling
To associate your repository with the fingerspelling topic, visit your repo's landing page and select "manage topics."