Gaze Tracking Based on Concatenating Spatial-Temporal Features

Based on experimental observations, there is a correlation between time and consecutive gaze positions in visual behaviors. Previous studies on gaze point estimation usually use images as the input for model trainings without taking into account the sequence relationship between image data. In addition to the spatial features, the temporal features are considered to improve the accuracy in this paper by using videos instead of images as the input data. To be able to capture spatial and temporal features at the same time, the convolutional neural network (CNN) and long short-term memory (LSTM) network are introduced to build a training model. In this way, CNN is used to extract the spatial features, and LSTM correlates temporal features. This paper presents a CNN Concatenating LSTM network (CCLN) that concatenates spatial and temporal features to improve the performance of gaze estimation in the case of time-series videos as the input training data. In addition, the proposed model can be optimized by exploring the numbers of LSTM layers, the influence of batch normalization (BN) and global average pooling layer (GAP) on CCLN. It is generally believed that larger amounts of training data will lead to better models. To provide data for training and prediction, we propose a method for constructing datasets of video for gaze point estimation. The issues are studied, including the effectiveness of different commonly used general models and the impact of transfer learning. Through exhaustive evaluation, it has been proved that the proposed method achieves a better prediction accuracy than the existing CNN-based methods. Finally, 93.1% of the best model and 92.6% of the general model MobileNet are obtained.

BJBN： BERT-JOIN-BiLSTM Networks for Medical Auxiliary Diagnostic

This study proposed a medicine auxiliary diagnosis model based on neural network. The model combines a bidirectional long short-term memory（Bi-LSTM）network and bidirectional encoder representations from transformers (BERT), which can well complete the extraction of local features of Chinese medicine texts. BERT can learn the global information of the text, so use BERT to get the global representation of medical text and then use Bi-LSTM to extract local features. We conducted a large number of comparative experiments on datasets. The results show that the proposed model has significant advantages over the state-of-the-art baseline model. The accuracy of the proposed model is 0.75.

Implementing Machine Learning Algorithms to Classify Postures and Forecast Motions When Using a Dynamic Chair

Many modern jobs require long periods of sitting on a chair that may result in serious health complications. Dynamic chairs are proposed as alternatives to the traditional sitting chairs; however, previous studies have suggested that most users are not aware of their postures and do not take advantage of the increased range of motion offered by the dynamic chairs. Building a system that identifies users’ postures in real time, as well as forecasts the next few postures, can bring awareness to the sitting behavior of each user. In this study, machine learning algorithms have been implemented to automatically classify users’ postures and forecast their next motions. The random forest, gradient decision tree, and support vector machine algorithms were used to classify postures. The evaluation of the trained classifiers indicated that they could successfully identify users’ postures with an accuracy above 90%. The algorithm can provide users with an accurate report of their sitting habits. A 1D-convolutional-LSTM network has also been implemented to forecast users’ future postures based on their previous motions, the model can forecast a user’s motions with high accuracy (97%). The ability of the algorithm to forecast future postures could be used to suggest alternative postures as needed.