# HAN Humanoid Object Interaction Model

This model is designed for humanoid robot interaction training.
It classifies object interaction behaviors such as grabbing,
pushing, pulling, carrying, and releasing objects.

## Model Type
Neural Network - Multi-class Classification

## Training Objective
To help humanoid robots recognize and categorize
basic object interaction behaviors.

## Input
Text-based action descriptions or structured motion signals

## Output
Interaction category label

## Categories
- grabbing
- pushing
- pulling
- carrying
- releasing

## Use Case
Robotics AI training
Simulation learning
Human-object interaction research

## License
MIT