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