ABHISHEK U K (STM21CS006)
ALBIN BINU (STM21CS015)
AYISHA ZOOMI (STM21CS022)
NASLA SAFIYA K (STM21CS045)

With the increasing need for contactless interactions in public and shared environments, the demand for hygienic and efficient input methods has risen significantly.This project introduces a novel contactless keyboard model that enables users to type by tracking and interpreting hand movements on a flat surface. Designed to enhance hygiene, accessibility and convenience. This system is ideal for scenarios where conventional keyboards are impractical.The system utilizes an external webcam to capture finger movements, using Mediapipe’s hand tracking module to detect hand landmarks. From the captured handlandmarks, key features including angles and distances between some hand joints are extracted, totaling 44 features. These 44 features are then used as inputs for a neural network model. This model, utilizing a Multi-Layer Perceptron (MLP) architecture, classifies 29 key classes: the 26 English letters, space, backspace and a static hand position for controlled pauses. Upon classification, the system uses the pyAutoGUI library to simulate keystrokes, providing visual feedback in real time to ensure an intuitive
user experience while reducing errors.
By eliminating physical contact, the system offers a seamless and efficient typing solution suited to modern computing needs. Improve accessibility by offering an alternative typing method for people with motor impairments. Its portability supports integration into immersive computing environments, including virtual and augmented reality, where physical keyboards may be impractical.

This innovative model introduces a practical and adaptable input method for hygiene-sensitive environments, public spaces and MR applications. Future improvements may include advanced machine learning techniques to enhance accuracy, expanded language
support and a refined user interface. This research highlights the transformative role of computer vision and artificial intelligence in redefining human-computer interaction, paving the way for a more hygienic, accessible and user-friendly typing solution.