Classification of Indoor Actions through Deep Neural Networks

Vella, Filippo; Augello, Agnese; Maniscalco, Umberto; Bentivenga, Vincenzo; Gaglio, Salvatore

doi:10.1109/sitis.2016.22

Cited by 3 publications

(4 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The experiment results show that the performance achieved by the proposed models using all 20 activities (class) is much higher than the performance obtained in [9] and [10], when the authors merged some classes as a strategy to create classes with more samples. They gave the merged classes more generic labels to avoid the problem of imbalance classes.…”

Section: Comparison With State-of-the-art Methodsmentioning

confidence: 82%

“…They gave the merged classes more generic labels to avoid the problem of imbalance classes. Vella et al [9] used CNN and the best results achieved is 76%, and Cipolla et al [10] used the LSTM and the best results achieved is 83.2%. A comparison results of the proposed methods and state-of-the-art methods specifically which were also applied to the SPHERE dataset are listed in TABLE XII.…”

Section: Comparison With State-of-the-art Methodsmentioning

confidence: 99%

“…Via their multi-layered architecture, DL methods can construct low level to high level representational features in sensor data hierarchically. Vella et al [9] used two different CNN architectures namely Deep Net 1 and Deep Net 2, and compared these with a baseline multilayer perceptron (MLP) using the SPHERE dataset. In this work, the authors merged some classes to macro labels to have classes with more samples and more generic labels as a strategy to dealing with imbalanced classes.…”

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

Classifying Imbalanced Multi-modal Sensor Data for Human Activity Recognition in a Smart Home using Deep Learning

Alani

Cosma

Taherkhani

2020

2020 International Joint Conference on Neural Networks (IJCNN)

View full text Add to dashboard Cite

In smart homes, data generated from real-time sensors for human activity recognition is complex, noisy and imbalanced. It is a significant challenge to create machine learning models that can classify activities which are not as commonly occurring as other activities. Machine learning models designed to classify imbalanced data are biased towards learning the more commonly occurring classes. Such learning bias occurs naturally, since the models better learn classes which contain more records. This paper examines whether fusing real-world imbalanced multi-modal sensor data improves classification results as opposed to using unimodal data; and compares deep learning approaches to dealing with imbalanced multi-modal sensor data when using various resampling methods and deep learning models. Experiments were carried out using a large multi-modal sensor dataset generated from the Sensor Platform for HEalthcare in a Residential Environment (SPHERE). The data comprises 16104 samples, where each sample comprises 5608 features and belongs to one of 20 activities (classes). Experimental results using SPHERE demonstrate the challenges of dealing with imbalanced multi-modal data and highlight the importance of having a suitable number of samples within each class for sufficiently training and testing deep learning models. Furthermore, the results revealed that when fusing the data and using the Synthetic Minority Oversampling Technique (SMOTE) to correct class imbalance, CNN-LSTM achieved the highest classification accuracy of 93.67% followed by CNN, 93.55%, and LSTM, i.e. 92.98%.

show abstract

Section: Comparison With State-of-the-art Methodsmentioning

confidence: 82%

Section: Comparison With State-of-the-art Methodsmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Classifying Imbalanced Multi-modal Sensor Data for Human Activity Recognition in a Smart Home using Deep Learning

Alani

Cosma

Taherkhani

2020

2020 International Joint Conference on Neural Networks (IJCNN)

View full text Add to dashboard Cite

show abstract

“…Deep learning has successfully been used for HAR [ 32 , 33 , 34 ]. Deep learning models have the ability to learn features in a hierarchical way starting from low level features reaching to high level ones.…”

Section: Related Workmentioning

confidence: 99%

Deep Learning of Fuzzy Weighted Multi-Resolution Depth Motion Maps with Spatial Feature Fusion for Action Recognition

et al. 2019

View full text Add to dashboard Cite

Human action recognition (HAR) is an important yet challenging task. This paper presents a novel method. First, fuzzy weight functions are used in computations of depth motion maps (DMMs). Multiple length motion information is also used. These features are referred to as fuzzy weighted multi-resolution DMMs (FWMDMMs). This formulation allows for various aspects of individual actions to be emphasized. It also helps to characterise the importance of the temporal dimension. This is important to help overcome, e.g., variations in time over which a single type of action might be performed. A deep convolutional neural network (CNN) motion model is created and trained to extract discriminative and compact features. Transfer learning is also used to extract spatial information from RGB and depth data using the AlexNet network. Different late fusion techniques are then investigated to fuse the deep motion model with the spatial network. The result is a spatial temporal HAR model. The developed approach is capable of recognising both human action and human–object interaction. Three public domain datasets are used to evaluate the proposed solution. The experimental results demonstrate the robustness of this approach compared with state-of-the art algorithms.

show abstract

Ambient intelligence-based multimodal human action recognition for autonomous systems

Jain

Gupta

et al. 2023

ISA Transactions

View full text Add to dashboard Cite

Classification of Indoor Actions through Deep Neural Networks

Cited by 3 publications

References 12 publications

Classifying Imbalanced Multi-modal Sensor Data for Human Activity Recognition in a Smart Home using Deep Learning

Classifying Imbalanced Multi-modal Sensor Data for Human Activity Recognition in a Smart Home using Deep Learning

Deep Learning of Fuzzy Weighted Multi-Resolution Depth Motion Maps with Spatial Feature Fusion for Action Recognition

Ambient intelligence-based multimodal human action recognition for autonomous systems

Contact Info

Product

Resources

About