A Hierarchical Approach to Activity Recognition and Fall Detection Using Wavelets and Adaptive Pooling

Human activity recognition has been a key study topic in the development of cyber physical systems and assisted living applications. In particular, inertial sensor based systems have become increasingly popular because they do not restrict users’ movement and are also relatively simple to implement compared to other approaches. In this paper, we present a hierarchical classification framework based on wavelets and adaptive pooling for activity recognition and fall detection predicting fall direction and severity. To accomplish this, windowed segments were extracted from each recording of inertial measurements from the SisFall dataset. A combination of wavelet based feature extraction and adaptive pooling was used before a classification framework was applied to determine the output class. Furthermore, tests were performed to determine the best observation window size and the sensor modality to use. Based on the experiments the best window size was found to be 3 s and the best sensor modality was found to be a combination of accelerometer and gyroscope measurements. These were used to perform activity recognition and fall detection with a resulting weighted F1 score of 94.67%. This framework is novel in terms of the approach to the human activity recognition and fall detection problem as it provides a scheme that is computationally less intensive while providing promising results and therefore can contribute to edge deployment of such systems.

[1]  Guang-Zhong Yang,et al.  Direction sensitive fall detection using a triaxial accelerometer and a barometric pressure sensor , 2011, 2011 Annual International Conference of the IEEE Engineering in Medicine and Biology Society.

[2]  Max A. Little,et al.  Freezing of gait and fall detection in Parkinson’s disease using wearable sensors: a systematic review , 2017, Journal of Neurology.

[3]  H. Afzal,et al.  NT-FDS—A Noise Tolerant Fall Detection System Using Deep Learning on Wearable Devices , 2021, Sensors.

[4]  J. Stevens,et al.  Medical Costs of Fatal and Nonfatal Falls in Older Adults , 2018, Journal of the American Geriatrics Society.

[5]  Md. Zahurul Islam,et al.  Real time direction-sensitive fall detection system using accelerometer and learning classifier , 2017, 2017 4th International Conference on Advances in Electrical Engineering (ICAEE).

[6]  Zhizeng Luo,et al.  Evaluation of Feature Extraction and Recognition for Activity Monitoring and Fall Detection Based on Wearable sEMG Sensors , 2017, Sensors.

[7]  Jung Keun Lee,et al.  Determination of Fall Direction Before Impact Using Support Vector Machine , 2015 .

[8]  Adnan Nadeem,et al.  A dataset build using wearable inertial measurement and ECG sensors for activity recognition, fall detection and basic heart anomaly detection system , 2019, Data in brief.

[9]  Muhammad Awais Azam,et al.  Activity-Aware Fall Detection and Recognition Based on Wearable Sensors , 2019, IEEE Sensors Journal.

[10]  Josef Börcsök,et al.  Monitoring System for Prediction and Detection of Epilepsy Seizure , 2019, 2019 Fourth International Conference on Advances in Computational Tools for Engineering Applications (ACTEA).

[11]  Eduardo Casilari-Pérez,et al.  A Study on the Application of Convolutional Neural Networks to Fall Detection Evaluated with Multiple Public Datasets , 2020, Sensors.

[12]  Marjorie Skubic,et al.  VAMPIR- an automatic fall detection system using a vertical PIR sensor array , 2012, 2012 6th International Conference on Pervasive Computing Technologies for Healthcare (PervasiveHealth) and Workshops.

[13]  Jesús Francisco Vargas-Bonilla,et al.  SisFall: A Fall and Movement Dataset , 2017, Sensors.

[14]  Salah Bourennane,et al.  A Tensor Approach for Activity Recognition and Fall Detection Using Wearable Inertial Sensors , 2020, 2020 First International Conference of Smart Systems and Emerging Technologies (SMARTTECH).

[15]  Xiao Zhang,et al.  Device-Free Simultaneous Wireless Localization and Activity Recognition With Wavelet Feature , 2017, IEEE Transactions on Vehicular Technology.

[16]  Senem Velipasalar,et al.  Automatic Fall Detection and Activity Classification by a Wearable Embedded Smart Camera , 2013, IEEE Journal on Emerging and Selected Topics in Circuits and Systems.

[17]  Jong-Hoon Youn,et al.  Fall Direction Detection using the Components of Acceleration Vector and Orientation Sensor on the Smartphone Environment , 2015 .

[18]  Jian Sun,et al.  Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition , 2014, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[19]  Rajasekar Mohan,et al.  A novel helmet design and implementation for drowsiness and fall detection of workers on-site using EEG and Random-Forest Classifier , 2019, ANT/EDI40.

[20]  C. Brayne,et al.  Inability to get up after falling, subsequent time on floor, and summoning help: prospective cohort study in people over 90 , 2008, BMJ : British Medical Journal.

[21]  Jongman Kim,et al.  Evaluation of Inertial Sensor-Based Pre-Impact Fall Detection Algorithms Using Public Dataset , 2019, Sensors.

[22]  Alex A. Freitas,et al.  A survey of hierarchical classification across different application domains , 2010, Data Mining and Knowledge Discovery.

[23]  Karin A Mack,et al.  Gender differences in seeking care for falls in the aged Medicare population. , 2012, American journal of preventive medicine.

[24]  Jean Meunier,et al.  Fall Detection from Human Shape and Motion History Using Video Surveillance , 2007, 21st International Conference on Advanced Information Networking and Applications Workshops (AINAW'07).

[25]  Luca Palmerini,et al.  Accelerometer-Based Fall Detection Using Machine Learning: Training and Testing on Real-World Falls , 2020, Sensors.

[26]  Md. Zahurul Islam,et al.  A direction-sensitive fall detection system using single 3D accelerometer and learning classifier , 2016, 2016 International Conference on Medical Engineering, Health Informatics and Technology (MediTec).

[27]  Adel Said Elmaghraby,et al.  Determining Fall direction and severity using SVMs , 2020, 2020 IEEE International Symposium on Signal Processing and Information Technology (ISSPIT).

[28]  Gabriella Balestra,et al.  Comparison of Different Sets of Features for Human Activity Recognition by Wearable Sensors , 2018, Sensors.

[29]  P. Bruegger,et al.  A Machine Learning Multi-Class Approach for Fall Detection Systems Based on Wearable Sensors with a Study on Sampling Rates Selection † , 2021, Sensors.

[30]  Janet Light,et al.  Optimizing pressure sensor array data for a smart-shoe fall monitoring system , 2015, 2015 IEEE SENSORS.

[31]  Bummo Koo,et al.  The Performance of Post-Fall Detection Using the Cross-Dataset: Feature Vectors, Classifiers and Processing Conditions , 2021, Sensors.

[32]  Zhuo Wang,et al.  Possible Life Saver: A Review on Human Fall Detection Technology , 2020, Robotics.

[33]  Bunthit Watanapa,et al.  Post-Fall Intelligence Supporting Fall Severity Diagnosis Using Kinect Sensor , 2018, Appl. Comput. Intell. Soft Comput..

[34]  Walid Gomaa,et al.  Competitive Feature Extraction for Activity Recognition based on Wavelet Transforms and Adaptive Pooling , 2019, 2019 International Joint Conference on Neural Networks (IJCNN).

[35]  Abbes Amira,et al.  Multiple comparator classifier framework for accelerometer-based fall detection and diagnostic , 2016, Appl. Soft Comput..

[36]  Kai-Chun Liu,et al.  Multiphase Identification Algorithm for Fall Recording Systems Using a Single Wearable Inertial Sensor , 2021, Sensors.

[37]  Vitoantonio Bevilacqua,et al.  Fall detection in indoor environment with kinect sensor , 2014, 2014 IEEE International Symposium on Innovations in Intelligent Systems and Applications (INISTA) Proceedings.