Predicting fall risk is necessary for rescue and accident prevention in the elderly. In this study, deep learning regression models were used to predict the acceleration sum vector magnitude (SVM) peak value, which represents the risk of a fall. Twenty healthy adults (aged 22.0±1.9 years, height 164.9±5.9 cm, weight 61.4±17.1 kg) provided data for 14 common daily life activities (ADL) and 11 falls using IMU (Inertial Measurement Unit) sensors (Movella Dot, Netherlands) at the S2. The input data includes information from 0.7 to 0.2 seconds before the acceleration SVM peak, encompassing 6-axis IMU data, as well as acceleration SVM and angular velocity SVM, resulting in a total of 8 feature vectors used to model training. Data augmentations were applied to solve data imbalances. The data was split into a 4 : 1 ratio for training and testing. The models were trained using Mean Squared Error (MSE) and Mean Absolute Error (MAE). The deep learning model utilized 1D-CNN and LSTM. The model with data augmentation exhibited lower error values in both MAE (1.19 g) and MSE (2.93g²). Low-height falls showed lower predicted acceleration peak values, while ADLs like jumping and sitting showed higher predicted values, indicating higher risks.

Deaf people use their own national sign or finger languages for communication. They have a lot of inconvenience in both social and financial problems. In this study, a finger language recognition system using an ensemble machine learning algorithm with an armband sensor of 8 channel surface electromyography (sEMG) is introduced. The algorithm consisted of signal acquisition, digital filtering, feature vector extraction, and an ensemble classifier based on artificial neural network (EANN). It was evaluated with Korean finger language (14 consonants, 17 vowels and 7 numbers) in 20 normal subjects. EANN was categorized with the number of classifiers (1 to 10) and the size of training data (50 to 1500). Mean accuracies and standard deviations for each structure were then obtained. Results showed that, as the number of classifiers (1 to 8) and the size of training data (50 to 300) were increased, the average accuracy of the E-ANN classifier was increased while the standard deviation was decreased. Statistical analysis showed that the optimal E-ANN structure was composed with 8 classifiers and 300 training data. This study suggested that E-ANN was more accurate than the general ANN for sign/finger language recognition.