Conclusion
Although the inference engine was not able to classify between left and right accurately, overall the performance was satisfying. Also, there were some cases where the gesture command needed to be applied multiple times to be able to detect and classify.
These issues could be better addressed by adding more training datasets and making the model more flexible.
What next !!
- Training the model with more test data for more accurate classification.
- Audio feedback depicting the gesture detected.
References
https://arc.aiaa.org/doi/10.2514/6.2018-4231
https://dl-cdn.ryzerobotics.com/downloads/Tello/Tello%20SDK%202.0%20User%20Guide.pdf
https://store.dji.com/product/tello
https://github.com/akshayvernekar/telloArduino
https://docs.m5stack.com/en/quick_start/m5core/m5stack_core_get_started_Arduino_Windows
https://www.arduino.cc/reference/en/libraries/m5stack/
https://shop.m5stack.com/products/fire-iot-development-kit?variant=16804798169178
https://docs.edgeimpulse.com/docs/continuous-motion-recognition
https://docs.edgeimpulse.com/docs/using-your-mobile-phone
Blog Series:
Gesture Controlled Drone #1 : Introduction
Gesture Controlled Drone #2 : Preperation - Drone
Gesture Controlled Drone #3 : Preperation - M5Stack
Gesture Controlled Drone #4 : Gesture Control
Gesture Controlled Drone #5 : Gesture Recognition using Edge Impulse