Skip to content

feat: update fixed weight decoder for improved performance and add tr…#25

Open
ArnavBhalla wants to merge 1 commit intosciencecorp:calvin/inferencefrom
ArnavBhalla:calvin/inference
Open

feat: update fixed weight decoder for improved performance and add tr…#25
ArnavBhalla wants to merge 1 commit intosciencecorp:calvin/inferencefrom
ArnavBhalla:calvin/inference

Conversation

@ArnavBhalla
Copy link
Copy Markdown

…aining data collection

  • Changed publish rate from 10 Hz to 30 Hz in fixed_weight_decoder.hpp for better responsiveness.
  • Added a new Python script (collect_training_data.py) to collect paired training data from the Synapse app.
  • Created easy_test.json and easy_train.json configurations for testing and training modes.
  • Implemented a BCI game (bci_game.py) for a 60-second selection task using joystick input.
  • Added a run script (run.sh) to facilitate launching the game in demo or device mode.
  • Developed a training script (train_decoder.py) to train a PyTorch model for joystick direction decoding and export it to ONNX format.

…aining data collection

- Changed publish rate from 10 Hz to 30 Hz in fixed_weight_decoder.hpp for better responsiveness.
- Added a new Python script (collect_training_data.py) to collect paired training data from the Synapse app.
- Created easy_test.json and easy_train.json configurations for testing and training modes.
- Implemented a BCI game (bci_game.py) for a 60-second selection task using joystick input.
- Added a run script (run.sh) to facilitate launching the game in demo or device mode.
- Developed a training script (train_decoder.py) to train a PyTorch model for joystick direction decoding and export it to ONNX format.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant