The Use of Transfer Learning for Activity Recognition in Instances of Heterogeneous Sensing

Transfer learning is a growing field that can address the variability of activity recognition problems by reusing the knowledge from previous experiences to recognise activities from different conditions, resulting in the leveraging of resources such as training and labelling efforts. Although integrating ubiquitous sensing technology and transfer learning seem promising, there are some research opportunities that, if addressed, could accelerate the development of activity recognition. This paper presents TL-FmRADLs; a framework that converges the feature fusion strategy with a teacher/learner approach over the active learning technique to automatise the self-training process of the learner models. Evaluation TL-FmRADLs is conducted over InSync; an open access dataset introduced for the first time in this paper. Results show promising effects towards mitigating the insufficiency of labelled data available by enabling the learner model to outperform the teacher’s performance.

Download Full-text

Empirical Study and Improvement on Deep Transfer Learning for Human Activity Recognition

Sensors ◽

10.3390/s19010057 ◽

2018 ◽

Vol 19 (1) ◽

pp. 57 ◽

Cited By ~ 7

Author(s):

Renjie Ding ◽

Xue Li ◽

Lanshun Nie ◽

Jiazhen Li ◽

Xiandong Si ◽

...

Keyword(s):

Deep Learning ◽

Empirical Study ◽

Transfer Learning ◽

Activity Recognition ◽

Human Activity ◽

Human Activity Recognition ◽

Sensor Data ◽

Accurate Identification ◽

First Time ◽

Deep Learning Model

Human activity recognition (HAR) based on sensor data is a significant problem in pervasive computing. In recent years, deep learning has become the dominating approach in this field, due to its high accuracy. However, it is difficult to make accurate identification for the activities of one individual using a model trained on data from other users. The decline on the accuracy of recognition restricts activity recognition in practice. At present, there is little research on the transferring of deep learning model in this field. This is the first time as we known, an empirical study was carried out on deep transfer learning between users with unlabeled data of target. We compared several widely-used algorithms and found that Maximum Mean Discrepancy (MMD) method is most suitable for HAR. We studied the distribution of features generated from sensor data. We improved the existing method from the aspect of features distribution with center loss and get better results. The observations and insights in this study have deepened the understanding of transfer learning in the activity recognition field and provided guidance for further research.

Download Full-text

Multistage transfer learning technique for classifying rare medical datasets

Journal of Ambient Intelligence and Humanized Computing ◽

10.1007/s12652-021-02989-1 ◽

2021 ◽

Author(s):

R. V. Aswiga ◽

R. Aishwarya ◽

A. P. Shanthi

Keyword(s):

Transfer Learning ◽

Learning Technique

Download Full-text

Personalized Human Activity Recognition Based on Integrated Wearable Sensor and Transfer Learning

Sensors ◽

10.3390/s21030885 ◽

2021 ◽

Vol 21 (3) ◽

pp. 885 ◽

Cited By ~ 1

Author(s):

Zhongzheng Fu ◽

Xinrun He ◽

Enkai Wang ◽

Jun Huo ◽

Jian Huang ◽

...

Keyword(s):

Transfer Learning ◽

Activity Recognition ◽

Human Activity ◽

Sensor Node ◽

Learning Algorithm ◽

Joint Probability ◽

Human Activity Recognition ◽

Sensor Technology ◽

Measurement Unit ◽

Wearable Sensor

Human activity recognition (HAR) based on the wearable device has attracted more attention from researchers with sensor technology development in recent years. However, personalized HAR requires high accuracy of recognition, while maintaining the model’s generalization capability is a major challenge in this field. This paper designed a compact wireless wearable sensor node, which combines an air pressure sensor and inertial measurement unit (IMU) to provide multi-modal information for HAR model training. To solve personalized recognition of user activities, we propose a new transfer learning algorithm, which is a joint probability domain adaptive method with improved pseudo-labels (IPL-JPDA). This method adds the improved pseudo-label strategy to the JPDA algorithm to avoid cumulative errors due to inaccurate initial pseudo-labels. In order to verify our equipment and method, we use the newly designed sensor node to collect seven daily activities of 7 subjects. Nine different HAR models are trained by traditional machine learning and transfer learning methods. The experimental results show that the multi-modal data improve the accuracy of the HAR system. The IPL-JPDA algorithm proposed in this paper has the best performance among five HAR models, and the average recognition accuracy of different subjects is 93.2%.

Download Full-text

Classification of Skin Disease from Skin images using Transfer Learning Technique

2020 4th International Conference on Electronics, Communication and Aerospace Technology (ICECA) ◽

10.1109/iceca49313.2020.9297567 ◽

2020 ◽

Author(s):

Honey Janoria ◽

Jasmine Minj ◽

Pooja Patre

Keyword(s):

Transfer Learning ◽

Skin Disease ◽

Learning Technique

Download Full-text

Human interaction recognition method based on parallel multi-feature fusion network

Intelligent Data Analysis ◽

10.3233/ida-205217 ◽

2021 ◽

Vol 25 (4) ◽

pp. 809-823

Author(s):

Qing Ye ◽

Haoxin Zhong ◽

Chang Qu ◽

Yongmei Zhang

Keyword(s):

Activity Recognition ◽

Feature Fusion ◽

Gaussian Model ◽

Research Direction ◽

Recognition Algorithm ◽

Human Interaction ◽

Amount Of Information ◽

Interaction Recognition ◽

Time Periods ◽

Human Interaction Recognition

Human activity recognition is a key technology in intelligent video surveillance and an important research direction in the field of computer vision. However, the complexity of human interaction features and the differences in motion characteristics at different time periods have always existed. In this paper, a human interaction recognition algorithm based on parallel multi-feature fusion network is proposed. First of all, in view of the different amount of information provided by the different time periods of action, an improved time-phased video down sampling method based on Gaussian model is proposed. Second, the Inception module uses different scale convolution kernels for feature extraction. It can improve network performance and reduce the amount of network parameters at the same time. The ResNet module mitigates degradation problem due to increased depth of neural networks and achieves higher classification accuracy. The amount of information provided in the motion video in different stages of motion time is also different. Therefore, we combine the advantages of the Inception network and ResNet to extract feature information, and then we integrate the extracted features. After the extracted features are merged, the training is continued to realize parallel connection of the multi-feature neural network. In this paper, experiments are carried out on the UT dataset. Compared with the traditional activity recognition algorithm, this method can accomplish the recognition tasks of six kinds of interactive actions in a better way, and its accuracy rate reaches 88.9%.

Download Full-text

Kernel Generalized Canonical Correlation and a New Feature Fusion Strategy

Lecture Notes in Computer Science - Artificial Intelligence and Security ◽

10.1007/978-3-030-24271-8_44 ◽

2019 ◽

pp. 488-500

Author(s):

Lina Wei ◽

Quansen Sun ◽

Xizhan Gao

Keyword(s):

Canonical Correlation ◽

Feature Fusion ◽

Fusion Strategy ◽

New Feature

Download Full-text

Study on Human Activity Recognition Using Semi-Supervised Active Transfer Learning

Sensors ◽

10.3390/s21082760 ◽

2021 ◽

Vol 21 (8) ◽

pp. 2760

Author(s):

Seungmin Oh ◽

Akm Ashiquzzaman ◽

Dongsu Lee ◽

Yeonggwang Kim ◽

Jinsul Kim

Keyword(s):

Deep Learning ◽

Transfer Learning ◽

Activity Recognition ◽

Human Activity ◽

Random Sampling ◽

Data Gathering ◽

Human Activity Recognition ◽

Learning Models ◽

Active Transfer ◽

Existing Data

In recent years, various studies have begun to use deep learning models to conduct research in the field of human activity recognition (HAR). However, there has been a severe lag in the absolute development of such models since training deep learning models require a lot of labeled data. In fields such as HAR, it is difficult to collect data and there are high costs and efforts involved in manual labeling. The existing methods rely heavily on manual data collection and proper labeling of the data, which is done by human administrators. This often results in the data gathering process often being slow and prone to human-biased labeling. To address these problems, we proposed a new solution for the existing data gathering methods by reducing the labeling tasks conducted on new data based by using the data learned through the semi-supervised active transfer learning method. This method achieved 95.9% performance while also reducing labeling compared to the random sampling or active transfer learning methods.

Download Full-text

Enhanced Sensing and Activity Recognition System Using IoT for Healthcare

International Journal of Information Communication Technologies and Human Development ◽

10.4018/ijicthd.2021040103 ◽

2021 ◽

Vol 13 (2) ◽

pp. 42-49

Author(s):

Yamini G. ◽

Gopinath Ganapathy

Keyword(s):

Activity Recognition ◽

Human Activity ◽

Healthcare Services ◽

Recognition System ◽

Physical Activities ◽

Video Data ◽

Large Coupling ◽

Sensing Technology ◽

Medical Benefits ◽

Primary Focus

Through the integration of advanced algorithms and smart sensing technology in healthcare services, huge medical benefits could be gained by the aged and sick people in determining their activity recognition. Human activity recognition (HAR) is still in the research for the past decades that promotes recognition of physical activities automatically. The main aim of HAR is to obtain and analyze the physical activities of a person, which could be promoted through several in-built sensors examined in the form of video data. Through this technique, necessary information could be obtained that also helps in preventing significant risks and also averts or alerts unfortunate events from happening. However, there is no particular categorization for human activity, and there is no description of the particular events to occur. The objective of this paper is to propose a healthcare information system based on IoT where enhancing activity recognition is the primary focus. Human activities are supposed to be diverse; it is necessary to choose appropriate sensors and the effective placement of those sensors in recognizing specific activities. One of the major challenges here is choosing the appropriate sensor for that particular instance and gathering data under particular circumstances. Due to the large coupling of sensors and their activity monitoring functionality, the solution to promote feasibility for the HAR predicament cannot be determined. A distinguishing feature of this paper is that it includes future users' perspectives.

Download Full-text