Real-Time Human Motion Tracking by Tello EDU Drone

Article Computer Science, Artificial Intelligence

RGB camera-based fallen person detection system embedded on a mobile platform

Sergio Lafuente-Arroyo et al.

Summary: This paper introduces a low-cost autonomous assistive patrol robot that can detect fallen persons and be deployed in care centers. By leveraging the YOLO network, the robot performs robust fallen person detection in various situations.

EXPERT SYSTEMS WITH APPLICATIONS (2022)

添加到收藏夹

Article Engineering, Civil

Enhanced Few-Shot Learning for Intrusion Detection in Railway Video Surveillance

Xiao Gong et al.

Summary: This paper proposes an efficient few-shot learning solution to address the challenging issue of railway intrusion detection, utilizing an enhanced model-agnostic meta-learner to improve accuracy. The proposed method is tested on a realistic railway video dataset, and the results show its superiority over standard supervised learning methods.

IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS (2022)

添加到收藏夹

Proceedings Paper Computer Science, Information Systems

VaIS: A Leading Visual and Inertial Dataset of Squats

Ahmed Fayez et al.

Summary: This paper presents a new Visual and Inertial dataset VaIS Dataset focusing on the analysis of squats exercises. The dataset includes videos and IMU data captured by different hardware systems, recording the squats performed by 24 males and 3 females.

PROCEEDINGS OF THE 2022 16TH INTERNATIONAL CONFERENCE ON UBIQUITOUS INFORMATION MANAGEMENT AND COMMUNICATION (IMCOM 2022) (2022)

添加到收藏夹

Article Chemistry, Analytical

ExerSense: Physical Exercise Recognition and Counting Algorithm from Wearables Robust to Positioning

Shun Ishii et al.

Summary: The study compared the accuracy of different wearable devices in recognizing multiple exercises, with the chest-mounted sensor performing the best and the ear-mounted sensor being the worst.

SENSORS (2021)

添加到收藏夹

Article Multidisciplinary Sciences

An eight-camera fall detection system using human fall pattern recognition via machine learning by a low-cost android box

Francy Shu et al.

Summary: The system is capable of detecting various types of falls at 60 meters and beyond, including through transparent materials. By analyzing the fall patterns and using machine learning, true falls can be differentiated from daily activities, and the system can be configured into different applications to capture every single fall event in high-risk areas.

SCIENTIFIC REPORTS (2021)

添加到收藏夹

Article Biotechnology & Applied Microbiology