YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
Quantum Humanoid Motion v3
Overview
Quantum Humanoid Motion v3 is an advanced BiLSTM + Transformer Decoder model built for humanoid robot dynamic motion understanding and prediction.
Model Details
- Architecture: BiLSTM + Transformer Decoder
- Framework: PyTorch
- Tasks: Action Recognition & Future Motion Generation
- Number of Classes: 10
Input Format
- Sequence Length: 100 frames
- Joints: 26
- Features per Joint: 6 (x, y, z, vx, vy, vz)
Output
- Action Label
- Confidence Score
- Predicted Future Frames
Supported Actions
- walking
- running
- jumping
- sitting
- standing
- climbing
- grabbing_object
- throwing
- turning_left
- turning_right
Hyperparameters
- Learning Rate: 0.00015
- Batch Size: 48
- Epochs: 75
- Optimizer: AdamW
- Dropout: 0.1
- Loss Function: CrossEntropy + MSE
Dataset
- 20,000 motion sequences
- 10 action classes
- Augmented with rotation, noise, and time-warping
License
MIT
- Downloads last month
- -
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support