GitHub Navigator: Your AI-Powered Repository Guide Using Pydantic AI
Dr.R.Poornima
AIML Dept
Assistant Professor
Poorniom
@gmail.com
M.Prudhvinath
CSE-AIML
MRUH Student
Prudhvinath07
@gmail.com
Aasmin Jainab
CSE-AIML
MRUH Student
Aasminjainab
@gmail.com
Vikas Chowdhary
CSE-AIML
MRUH Student
Vikassirvi9492
@gmail.com
K.Ragha Sathwika
CSE-AIML
MRUH Student
Raghasathwika
kondapalli@gmail.com
K.Raghavendra
CSE-AIML
MRUH Student
Karkaraghavendra97
@gmail.com
ABSTRACT : The project was entitled "Real-Time Fingerspelling Recognition Using MediaPipe and Adaptive Neural Networks". Real-Time Fingerspelling Recognition Using MediaPipe and Adaptive Neural Networks is designed to facilitate communication by translating Sign Language fingerspelling into both written text and spoken words. The system addresses a crucial accessibility challenge faced by the deaf and hard-of-hearing community. A computer vision-based approach is adopted, utilizing a readily available webcam as the primary sensory input. Hand motions are captured, and the MediaPipe toolkit is employed to reliably detect hands and extract key anatomical landmarks. The system then generates a skeletal representation of the hand, effectively circumventing issues stemming from fluctuating backgrounds and variable lighting. These extracted skeletal data points become the input for a specially constructed Convolutional Neural Network (CNN). When a gesture is successfully classified, its equivalent text is displayed, and concurrently transformed into speech via the pyttsx3 module, providing a multi-faceted output. Additionally, the system integrates word suggestions using the enchant library, enhancing communication fluidity and enabling user-led corrections via the interface. A user-friendly interface, built using Tkinter and Pillow, offers real-time visual feedback on hand tracking, gesture identification, alongside text and audio outputs. The project aims to provide an essential and cost-effective means of breaking down communication barriers between the hearing and deaf communities, with the aim of fostering greater inclusion and simplified interaction.
Keywords : GitHub API, Repository Analysis, AI-Powered Tools, Pydantic, Natural Language Querying, Software Engineering, Code Quality, Data Retrieval, Open-Source
