Activity Detection in Indoor Environments Using Multiple 2D Lidars.

In health monitoring systems for the elderly, a crucial aspect is unobtrusively and continuously monitoring their activities to detect potentially hazardous incidents such as sudden falls as soon as they occur. However, the effectiveness of current non-contact sensor-based activity detection systems is limited by obstacles present in the environment. To overcome this limitation, a straightforward yet highly efficient approach involves utilizing multiple sensors that collaborate seamlessly. This paper proposes a method that leverages 2D Light Detection and Ranging (Lidar) technology for activity detection. Multiple 2D Lidars are positioned in an indoor environment with varying obstacles such as furniture, working cohesively to create a comprehensive representation of ongoing activities. The data from these Lidars is concatenated and transformed into a more interpretable format, resembling images. A convolutional Long Short-Term Memory (LSTM) Neural Network is then used to process these generated images to classify the activities. The proposed approach achieves high accuracy in three tasks: activity detection, fall detection, and unsteady gait detection. Specifically, it attains accuracies of 96.10%, 99.13%, and 93.13% for these tasks, respectively. This demonstrates the efficacy and promise of the method in effectively monitoring and identifying potentially hazardous events for the elderly through 2D Lidars, which are non-intrusive sensing technology.

Facial Action Unit Detection using 3D Face Landmarks for Pain Detection.

Automatic detection of facial action units (AUs) has recently gained attention for its applications in facial expression analysis. However, using AUs in research can be challenging since they are typically manually annotated, which can be time-consuming, repetitive, and error-prone. Advancements in automated AU detection can greatly reduce the time required for this task and improve the reliability of annotations for downstream tasks, such as pain detection. In this study, we present an efficient method for detecting AUs using only 3D face landmarks. Using the detected AUs, we trained state-of-the-art deep learning models to detect pain, which validates the effectiveness of the AU detection model. Our study also establishes a new benchmark for pain detection on the BP4D+ dataset, demonstrating an 11.13% improvement in F1-score and a 3.09% improvement in accuracy using a Transformer model compared to existing studies. Our results show that utilizing only eight predicted AUs still achieves competitive results when compared to using all 34 ground-truth AUs.