This technology is an innovative recursive probabilistic neural network for accurate identification of brushing posture and position, and the correctness and integrity of brushing according to Bayesian brushing method. The advantages of algorithms with high accuracy and low computational complexity provide forward-looking and product-oriented technologies for related industries.

The technology is based on the Recurrent Probabilistic Neural Network smart toothbrush. The recognition rate is up to 98.64%, which is 16.2% more than the CNN model and 21.21% more than the LSTM model. This model is extremely small, enabling instant identification in low-cost embedded systems, reducing the cost of smart toothbrushes and improving recognition accuracy and speed.

This technology proposes a high-performance recursive probabilistic neural network, which is applied to the brushing posture recognition, improves the evaluation accuracy of the Bayesian brushing method, increases the product value, the best recognition rate and less computing resource demand, and is low-cost and low-power. Consumption of consumer electronics.