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