The problem Cogni-Glove solves

It is primarily for differently abled peoples to make tech more accessible to them. In future it can help in health care as well by implementing biomedical sensors to enhance it's functionality. By using this device they won't have to dependable on others for their need. For vocally disabled peoples, it can remove communication gap by translating sign languages.

In the current interation it can detect pretrained gestures to perform actions. We have made a web app to show the detected gestures in a dashboard.

Future goals

• Complete finger and gesture detection for complete sign language translation. (also support for different sign languages).
• Use of hall effect sensors to remove any strain experienced on fingers
• Improved MPU for complex gesture detection and on device detection (edge AI aplication).
• In-build mic to communicate with AI agents over network.

Challenges we ran into

Problems faced:

• Diffulcity in creation of training data collection. Created a simple GUI application in python using trinket
• Gesture conflict due to low sample rate. Solved by increasing sampling rate
• Unable to incorporate all envisioned features due to low program memory space. Solved by removing Bluetooth HID feature.
• Double clicking issue due to no debouncing.
• Websocket implementation for webapp state updation.