Add build doc.
Add document outlining build steps.
Update portal wake.
After sleeping through the evening have the portal reload.
Remove unused examples.
Cleaning up demos I used functionality from to confirm ESP EYE behavior but are not part of the core project.
Fixed wifi initialization. The application now recognizes the wake word visual, and then performs the correct PUT request when hearing on/off.
Fix the conditional for when to deep sleep.
Update cert doc and tags.
Add docstring for how the cert is generated for future use, and move the char const.
Updated board to initialize wifi at start.
Word recognition update.
Remove extra feature spectrograms. Update how command response is handled and the loop for determining if the device is listening.
Model working with keywords.
Model trained to respond to visual, on and off. Training scripts and custom features updated for full run.
Updates from training.
Setting updates after using all to go through a full training setup.
Micro speech example running on the ESP-EYE with some mods.
Model training scripts.
Add scripts for training data generation, model training and artifact generation.
Remove top level files moved into sig previously.
After running into issues with TF demo get ESP WHO working on ESP-EYE.
Add ESP http post.
Verify ESP HTTP POST functionality to the cloudflare listener.
Update to shut off display from 9PM to 6AM EST.
From time to time the ESP32 timesout or runs into an error. Catch and use supervisor to reload. This loop is read only so no data/integrity issues from the crash.
Initial ESP32 code.
Code to signal the service endpoint from the ESP32 inside of a FreeRTOS task.
Update sighandler toggle.
Worker functions more like a toggle now instead of doing various checks and parameter parsing.