suixin1424/mouse_control
一种基于神经网络来模拟人手移动鼠标的方法
Uses a shallow 3-layer fully-connected neural network trained on collected human mouse trajectories to generate naturalistic cursor paths, outputting 10 interpolation points from displacement inputs (dx, dy). Includes data collection via interactive GUI, model training pipeline, and ONNX export for inference in C++ and Python environments. Addresses limitations of traditional trajectory methods (PID controllers) by learning human-specific movement patterns from real cursor data.
190 stars. No commits in the last 6 months.
Stars
190
Forks
50
Language
Python
License
—
Category
Last pushed
Jan 21, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/suixin1424/mouse_control"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related frameworks
singharyans754/AI-Virtual-Mouse
An AI-based application which let us control the mouse cursor by moving finger in Air .
SunOner/mouse_net
Mouse Movement Prediction Training Tool
MatheusR42/yeelight-hand-controller
With Yeelight Hand Controller you can turn on and off your Yeelight just with a finger snap and...
VitorCarvalho67/Barium
Cutting-edge Python software utilizing neural networks and machine learning for precise hand...
omana1/Motion-controlled-3D-game
🍉 3D Fruit Slicer Game with PoseNet & Three.js An experimental browser-based game that uses...