scholarly journals Data Augmentation vs. Domain Adaptation—A Case Study in Human Activity Recognition

Technologies ◽  
2020 ◽  
Vol 8 (4) ◽  
pp. 55
Author(s):  
Evaggelos Spyrou ◽  
Eirini Mathe ◽  
Georgios Pikramenos ◽  
Konstantinos Kechagias ◽  
Phivos Mylonas
Keyword(s):  
Activity Recognition ◽  
Human Activity ◽  
Data Augmentation ◽  
Domain Adaptation ◽  
Recognition Task ◽  
Human Activity Recognition ◽  
Video Data ◽  
Training Dataset ◽  
Data Systems ◽  
Big Data Systems

Recent advances in big data systems and databases have made it possible to gather raw unlabeled data at unprecedented rates. However, labeling such data constitutes a costly and timely process. This is especially true for video data, and in particular for human activity recognition (HAR) tasks. For this reason, methods for reducing the need of labeled data for HAR applications have drawn significant attention from the research community. In particular, two popular approaches developed to address the above issue are data augmentation and domain adaptation. The former attempts to leverage problem-specific, hand-crafted data synthesizers to augment the training dataset with artificial labeled data instances. The latter attempts to extract knowledge from distinct but related supervised learning tasks for which labeled data is more abundant than the problem at hand. Both methods have been extensively studied and used successfully on various tasks, but a comprehensive comparison of the two has not been carried out in the context of video data HAR. In this work, we fill this gap by providing ample experimental results comparing data augmentation and domain adaptation techniques on a cross-viewpoint, human activity recognition task from pose information.

AdaptNet: Human Activity Recognition via Bilateral Domain Adaptation using Semi-Supervised Deep Translation Networks

2021 ◽  
pp. 1-1
Author(s):  
Sungtae An ◽  
Alessio Medda ◽  
Michael N. Sawka ◽  
Clayton J. Hutto ◽  
Mindy L. Millard-Stafford ◽  
...  
Keyword(s):  
Activity Recognition ◽  
Human Activity ◽  
Domain Adaptation ◽  
Human Activity Recognition

scholarly journals Human Activity Recognition Using Binary Sensors, BLE Beacons, an Intelligent Floor and Acceleration Data: A Machine Learning Approach

Proceedings ◽  
2018 ◽  
Vol 2 (19) ◽  
pp. 1265 ◽  
Author(s):  
Jesús D. Cerón ◽  
Diego M. López ◽  
Bjoern M. Eskofier
Keyword(s):  
Machine Learning ◽  
Activity Recognition ◽  
Human Activity ◽  
Class Imbalance ◽  
Human Activity Recognition ◽  
Training Dataset ◽  
Desktop Application ◽  
Acceleration Sensors ◽  
Machine Learning Approach ◽  
Mining Projects

Although there have been many studies aimed at the field of Human Activity Recognition, the relationship between what we do and where we do it has been little explored in this field. The objective of this paper is to propose an approach based on machine learning to address the challenge of the 1st UCAmI cup, which is the recognition of 24 activities of daily living using a dataset that allows to explore the aforementioned relationship, since it contains data collected from four data sources: binary sensors, an intelligent floor, proximity and acceleration sensors. The methodology for data mining projects CRISP-DM was followed in this work. To perform synchronization and classification tasks a java desktop application was developed. As a result, the accuracy achieved in the classification of the 24 activities using 10-fold-cross-validation on the training dataset was 92.1%, but an accuracy of 60.1% was obtained on the test dataset. The low accuracy of the classification might be caused by the class imbalance of the training dataset; therefore, more labeled data are necessary for training the algorithm. Although we could not obtain an optimal result, it is possible to iterate in the methodology to look for a way to improve the obtained results.

scholarly journals Complex Deep Neural Networks from Large Scale Virtual IMU Data for Effective Human Activity Recognition Using Wearables

Sensors ◽  
2021 ◽  
Vol 21 (24) ◽  
pp. 8337
Author(s):  
Hyeokhyen Kwon ◽  
Gregory D. Abowd ◽  
Thomas Plötz
Keyword(s):  
Activity Recognition ◽  
Human Activity ◽  
Large Scale ◽  
State Of The Art ◽  
Human Activity Recognition ◽  
The State ◽  
Model Parameters ◽  
Data Systems ◽  
Performance Improvements ◽  
Measurement Units

Supervised training of human activity recognition (HAR) systems based on body-worn inertial measurement units (IMUs) is often constrained by the typically rather small amounts of labeled sample data. Systems like IMUTube have been introduced that employ cross-modality transfer approaches to convert videos of activities of interest into virtual IMU data. We demonstrate for the first time how such large-scale virtual IMU datasets can be used to train HAR systems that are substantially more complex than the state-of-the-art. Complexity is thereby represented by the number of model parameters that can be trained robustly. Our models contain components that are dedicated to capture the essentials of IMU data as they are of relevance for activity recognition, which increased the number of trainable parameters by a factor of 1100 compared to state-of-the-art model architectures. We evaluate the new model architecture on the challenging task of analyzing free-weight gym exercises, specifically on classifying 13 dumbbell execises. We have collected around 41 h of virtual IMU data using IMUTube from exercise videos available from YouTube. The proposed model is trained with the large amount of virtual IMU data and calibrated with a mere 36 min of real IMU data. The trained model was evaluated on a real IMU dataset and we demonstrate the substantial performance improvements of 20% absolute F1 score compared to the state-of-the-art convolutional models in HAR.

scholarly journals An Information Gain-Based Model and an Attention-Based RNN for Wearable Human Activity Recognition

Entropy ◽  
2021 ◽  
Vol 23 (12) ◽  
pp. 1635
Author(s):  
Leyuan Liu ◽  
Jian He ◽  
Keyan Ren ◽  
Jonathan Lungu ◽  
Yibin Hou ◽  
...  
Keyword(s):  
Activity Recognition ◽  
Human Activity ◽  
Short Term Memory ◽  
Information Gain ◽  
Recognition Task ◽  
Human Activity Recognition ◽  
Short Term ◽  
Activity Model ◽  
Long Short Term Memory ◽  
Selection Of

Wearable sensor-based HAR (human activity recognition) is a popular human activity perception method. However, due to the lack of a unified human activity model, the number and positions of sensors in the existing wearable HAR systems are not the same, which affects the promotion and application. In this paper, an information gain-based human activity model is established, and an attention-based recurrent neural network (namely Attention-RNN) for human activity recognition is designed. Besides, the attention-RNN, which combines bidirectional long short-term memory (BiLSTM) with attention mechanism, was tested on the UCI opportunity challenge dataset. Experiments prove that the proposed human activity model provides guidance for the deployment location of sensors and provides a basis for the selection of the number of sensors, which can reduce the number of sensors used to achieve the same classification effect. In addition, experiments show that the proposed Attention-RNN achieves F1 scores of 0.898 and 0.911 in the ML (Modes of Locomotion) task and GR (Gesture Recognition) task, respectively.

scholarly journals Combining Supervised and Unsupervised Learning Algorithms for Human Activity Recognition

Sensors ◽  
2021 ◽  
Vol 21 (18) ◽  
pp. 6309
Author(s):  
Elena-Alexandra Budisteanu ◽  
Irina Georgiana Mocanu
Keyword(s):  
Unsupervised Learning ◽  
Activity Recognition ◽  
Human Activity ◽  
Data Augmentation ◽  
Learning Algorithms ◽  
Human Activity Recognition ◽  
Convolutional Networks ◽  
Supervised And Unsupervised Learning ◽  
Novel Approach ◽  
Unsupervised Deep Learning

Human activity recognition is an extensively researched topic in the last decade. Recent methods employ supervised and unsupervised deep learning techniques in which spatial and temporal dependency is modeled. This paper proposes a novel approach for human activity recognition using skeleton data. The method combines supervised and unsupervised learning algorithms in order to provide qualitative results and performance in real time. The proposed method involves a two-stage framework: the first stage applies an unsupervised clustering technique to group up activities based on their similarity, while the second stage classifies data assigned to each group using graph convolutional networks. Different clustering techniques and data augmentation strategies are explored for improving the training process. The results were compared against the state of the art methods and the proposed model achieved 90.22% Top-1 accuracy performance for NTU-RGB+D dataset (the performance was increased by approximately 9% compared with the baseline graph convolutional method). Moreover, inference time and total number of parameters stay within the same magnitude order. Extending the initial set of activities with additional classes is fast and robust, since there is no required retraining of the entire architecture but only to retrain the cluster to which the activity is assigned.

Sign in / Sign up

Export Citation Format

Share Document