Research

13 pages, 3233 KiB

Open AccessArticle

A Blanket Accommodative Sleep Posture Classification System Using an Infrared Depth Camera: A Deep Learning Approach with Synthetic Augmentation of Blanket Conditions

by Andy Yiu-Chau Tam, Bryan Pak-Hei So, Tim Tin-Chun Chan, Alyssa Ka-Yan Cheung, Duo Wai-Chi Wong and James Chung-Wai Cheung

Sensors 2021, 21(16), 5553; https://doi.org/10.3390/s21165553 - 18 Aug 2021

Cited by 20 | Viewed by 4602

Abstract

Surveillance of sleeping posture is essential for bed-ridden patients or individuals at-risk of falling out of bed. Existing sleep posture monitoring and classification systems may not be able to accommodate the covering of a blanket, which represents a barrier to conducting pragmatic studies. [...] Read more.

Surveillance of sleeping posture is essential for bed-ridden patients or individuals at-risk of falling out of bed. Existing sleep posture monitoring and classification systems may not be able to accommodate the covering of a blanket, which represents a barrier to conducting pragmatic studies. The objective of this study was to develop an unobtrusive sleep posture classification that could accommodate the use of a blanket. The system uses an infrared depth camera for data acquisition and a convolutional neural network to classify sleeping postures. We recruited 66 participants (40 men and 26 women) to perform seven major sleeping postures (supine, prone (head left and right), log (left and right) and fetal (left and right)) under four blanket conditions (thick, medium, thin, and no blanket). Data augmentation was conducted by affine transformation and data fusion, generating additional blanket conditions with the original dataset. Coarse-grained (four-posture) and fine-grained (seven-posture) classifiers were trained using two fully connected network layers. For the coarse classification, the log and fetal postures were merged into a side-lying class and the prone class (head left and right) was pooled. The results show a drop of overall F1-score by 8.2% when switching to the fine-grained classifier. In addition, compared to no blanket, a thick blanket reduced the overall F1-scores by 3.5% and 8.9% for the coarse- and fine-grained classifiers, respectively; meanwhile, the lowest performance was seen in classifying the log (right) posture under a thick blanket, with an F1-score of 72.0%. In conclusion, we developed a system that can classify seven types of common sleeping postures under blankets and achieved an F1-score of 88.9%. Full article

(This article belongs to the Special Issue Deep Learning Image Recognition Systems)

► Show Figures

Figure 1

19 pages, 23379 KiB

Open AccessArticle

Fast and Accurate Object Detection in Remote Sensing Images Based on Lightweight Deep Neural Network

by Lei Lang, Ke Xu, Qian Zhang and Dong Wang

Sensors 2021, 21(16), 5460; https://doi.org/10.3390/s21165460 - 13 Aug 2021

Cited by 17 | Viewed by 3252

Abstract

Deep learning-based object detection in remote sensing images is an important yet challenging task due to a series of difficulties, such as complex geometry scene, dense target quantity, and large variant in object distributions and scales. Moreover, algorithm designers also have to make [...] Read more.

Deep learning-based object detection in remote sensing images is an important yet challenging task due to a series of difficulties, such as complex geometry scene, dense target quantity, and large variant in object distributions and scales. Moreover, algorithm designers also have to make a trade-off between model’s complexity and accuracy to meet the real-world deployment requirements. To deal with these challenges, we proposed a lightweight YOLO-like object detector with the ability to detect objects in remote sensing images with high speed and high accuracy. The detector is constructed with efficient channel attention layers to improve the channel information sensitivity. Differential evolution was also developed to automatically find the optimal anchor configurations to address issue of large variant in object scales. Comprehensive experiment results show that the proposed network outperforms state-of-the-art lightweight models by 5.13% and 3.58% in accuracy on the RSOD and DIOR dataset, respectively. The deployed model on an NVIDIA Jetson Xavier NX embedded board can achieve a detection speed of 58 FPS with less than 10W power consumption, which makes the proposed detector very suitable for low-cost low-power remote sensing application scenarios. Full article

(This article belongs to the Special Issue Deep Learning Image Recognition Systems)

► Show Figures

Figure 1

13 pages, 44979 KiB

Open AccessArticle

GMSRI: A Texture-Based Martian Surface Rock Image Dataset

by Cong Wang, Zian Zhang, Yongqiang Zhang, Rui Tian and Mingli Ding

Sensors 2021, 21(16), 5410; https://doi.org/10.3390/s21165410 - 10 Aug 2021

Cited by 3 | Viewed by 2351

Abstract

CNN-based Martian rock image processing has attracted much attention in Mars missions lately, since it can help planetary rover autonomously recognize and collect high value science targets. However, due to the difficulty of Martian rock image acquisition, the accuracy of the processing model [...] Read more.

CNN-based Martian rock image processing has attracted much attention in Mars missions lately, since it can help planetary rover autonomously recognize and collect high value science targets. However, due to the difficulty of Martian rock image acquisition, the accuracy of the processing model is affected. In this paper, we introduce a new dataset called “GMSRI” that is a mixture of real Mars images and synthetic counterparts which are generated by GAN. GMSRI aims to provide a set of Martian rock images sorted by the texture and spatial structure of rocks. This paper offers a detailed analysis of GMSRI in its current state: Five sub-trees with 28 leaf nodes and 30,000 images in total. We show that GMSRI is much larger in scale and diversity than the current same kinds of datasets. Constructing such a database is a challenging task, and we describe the data collection, selection and generation processes carefully in this paper. Moreover, we evaluate the effectiveness of the GMSRI by an image super-resolution task. We hope that the scale, diversity and hierarchical structure of GMSRI can offer opportunities to researchers in the Mars exploration community and beyond. Full article

(This article belongs to the Special Issue Deep Learning Image Recognition Systems)

► Show Figures

Figure 1

21 pages, 51271 KiB

Open AccessArticle

Adversarial Attack and Defence through Adversarial Training and Feature Fusion for Diabetic Retinopathy Recognition

by Sheeba Lal, Saeed Ur Rehman, Jamal Hussain Shah, Talha Meraj, Hafiz Tayyab Rauf, Robertas Damaševičius, Mazin Abed Mohammed and Karrar Hameed Abdulkareem

Sensors 2021, 21(11), 3922; https://doi.org/10.3390/s21113922 - 07 Jun 2021

Cited by 66 | Viewed by 6104

Abstract

Due to the rapid growth in artificial intelligence (AI) and deep learning (DL) approaches, the security and robustness of the deployed algorithms need to be guaranteed. The security susceptibility of the DL algorithms to adversarial examples has been widely acknowledged. The artificially created [...] Read more.

Due to the rapid growth in artificial intelligence (AI) and deep learning (DL) approaches, the security and robustness of the deployed algorithms need to be guaranteed. The security susceptibility of the DL algorithms to adversarial examples has been widely acknowledged. The artificially created examples will lead to different instances negatively identified by the DL models that are humanly considered benign. Practical application in actual physical scenarios with adversarial threats shows their features. Thus, adversarial attacks and defense, including machine learning and its reliability, have drawn growing interest and, in recent years, has been a hot topic of research. We introduce a framework that provides a defensive model against the adversarial speckle-noise attack, the adversarial training, and a feature fusion strategy, which preserves the classification with correct labelling. We evaluate and analyze the adversarial attacks and defenses on the retinal fundus images for the Diabetic Retinopathy recognition problem, which is considered a state-of-the-art endeavor. Results obtained on the retinal fundus images, which are prone to adversarial attacks, are 99% accurate and prove that the proposed defensive model is robust. Full article

(This article belongs to the Special Issue Deep Learning Image Recognition Systems)

► Show Figures

Figure 1

16 pages, 2468 KiB

Open AccessArticle

LeanNet: An Efficient Convolutional Neural Network for Digital Number Recognition in Industrial Products

by Na Qin, Longkai Liu, Deqing Huang, Bi Wu and Zonghong Zhang

Sensors 2021, 21(11), 3620; https://doi.org/10.3390/s21113620 - 22 May 2021

Cited by 3 | Viewed by 2186

Abstract

The remarkable success of convolutional neural networks (CNNs) in computer vision tasks is shown in large-scale datasets and high-performance computing platforms. However, it is infeasible to deploy large CNNs on resource constrained platforms, such as embedded devices, on account of the huge overhead. [...] Read more.

The remarkable success of convolutional neural networks (CNNs) in computer vision tasks is shown in large-scale datasets and high-performance computing platforms. However, it is infeasible to deploy large CNNs on resource constrained platforms, such as embedded devices, on account of the huge overhead. To recognize the label numbers of industrial black material product and deploy deep CNNs in real-world applications, this research uses an efficient method to simultaneously (a) reduce the network model size and (b) lower the amount of calculation without compromising accuracy. More specifically, the method is implemented by pruning channels and corresponding filters that are identified as having a trivial effect on the output accuracy. In this paper, we prune VGG-16 to obtain a compact network called LeanNet, which gives a 25× reduction in model size and a 4.5× reduction in float point operations (FLOPs), while the accuracy on our dataset is close to the original accuracy by retraining the network. Besides, we also find that LeanNet could achieve better performance on reductions in model size and computation compared to some lightweight networks like MobileNet and SqueezeNet, which are widely used in engineering applications. This research has good application value in the field of industrial production. Full article

(This article belongs to the Special Issue Deep Learning Image Recognition Systems)

► Show Figures

Figure 1

14 pages, 4788 KiB

Open AccessArticle

Multi-U-Net: Residual Module under Multisensory Field and Attention Mechanism Based Optimized U-Net for VHR Image Semantic Segmentation

by Si Ran, Jianli Ding, Bohua Liu, Xiangyu Ge and Guolin Ma

Sensors 2021, 21(5), 1794; https://doi.org/10.3390/s21051794 - 05 Mar 2021

Cited by 6 | Viewed by 1978

Abstract

As the acquisition of very high resolution (VHR) images becomes easier, the complex characteristics of VHR images pose new challenges to traditional machine learning semantic segmentation methods. As an excellent convolutional neural network (CNN) structure, U-Net does not require manual intervention, and its [...] Read more.

As the acquisition of very high resolution (VHR) images becomes easier, the complex characteristics of VHR images pose new challenges to traditional machine learning semantic segmentation methods. As an excellent convolutional neural network (CNN) structure, U-Net does not require manual intervention, and its high-precision features are widely used in image interpretation. However, as an end-to-end fully convolutional network, U-Net has not explored enough information from the full scale, and there is still room for improvement. In this study, we constructed an effective network module: residual module under a multisensory field (RMMF) to extract multiscale features of target and an attention mechanism to optimize feature information. RMMF uses parallel convolutional layers to learn features of different scales in the network and adds shortcut connections between stacked layers to construct residual blocks, combining low-level detailed information with high-level semantic information. RMMF is universal and extensible. The convolutional layer in the U-Net network is replaced with RMMF to improve the network structure. Additionally, the multiscale convolutional network was tested using RMMF on the Gaofen-2 data set and Potsdam data sets. Experiments show that compared to other technologies, this method has better performance in airborne and spaceborne images. Full article

(This article belongs to the Special Issue Deep Learning Image Recognition Systems)

► Show Figures

Figure 1

14 pages, 3548 KiB

Open AccessArticle

Automated Counting Grains on the Rice Panicle Based on Deep Learning Method

by Ruoling Deng, Ming Tao, Xunan Huang, Kemoh Bangura, Qian Jiang, Yu Jiang and Long Qi

Sensors 2021, 21(1), 281; https://doi.org/10.3390/s21010281 - 04 Jan 2021

Cited by 16 | Viewed by 4363

Abstract

Grain number per rice panicle, which directly determines grain yield, is an important agronomic trait for rice breeding and yield-related research. However, manually counting grains of rice per panicle is time-consuming, laborious, and error-prone. In this research, a grain detection model was proposed [...] Read more.

Grain number per rice panicle, which directly determines grain yield, is an important agronomic trait for rice breeding and yield-related research. However, manually counting grains of rice per panicle is time-consuming, laborious, and error-prone. In this research, a grain detection model was proposed to automatically recognize and count grains on primary branches of a rice panicle. The model used image analysis based on deep learning convolutional neural network (CNN), by integrating the feature pyramid network (FPN) into the faster R-CNN network. The performance of the grain detection model was compared to that of the original faster R-CNN model and the SSD model, and it was found that the grain detection model was more reliable and accurate. The accuracy of the grain detection model was not affected by the lighting condition in which images of rice primary branches were taken. The model worked well for all rice branches with various numbers of grains. Through applying the grain detection model to images of fresh and dry branches, it was found that the model performance was not affected by the grain moisture conditions. The overall accuracy of the grain detection model was 99.4%. Results demonstrated that the model was accurate, reliable, and suitable for detecting grains of rice panicles with various conditions. Full article

(This article belongs to the Special Issue Deep Learning Image Recognition Systems)

► Show Figures

Figure 1

22 pages, 10496 KiB

Open AccessArticle

Quantitative Analysis of Metallographic Image Using Attention-Aware Deep Neural Networks

by Yifei Xu, Yuewan Zhang, Meizi Zhang, Mian Wang, Wujiang Xu, Chaoyong Wang, Yan Sun and Pingping Wei

Sensors 2021, 21(1), 43; https://doi.org/10.3390/s21010043 - 23 Dec 2020

Cited by 4 | Viewed by 2816

Abstract

As a detection tool to identify metal or alloy, metallographic quantitative analysis has received increasing attention for its ability to evaluate quality control and reveal mechanical properties. The detection procedure is mainly operated manually to locate and characterize the constitution in metallographic images. [...] Read more.

As a detection tool to identify metal or alloy, metallographic quantitative analysis has received increasing attention for its ability to evaluate quality control and reveal mechanical properties. The detection procedure is mainly operated manually to locate and characterize the constitution in metallographic images. The automatic detection is still a challenge even with the emergence of several excellent models. Benefiting from the development of deep learning, with regard to two different metallurgical structural steel image datasets, we propose two attention-aware deep neural networks, Modified Attention U-Net (MAUNet) and Self-adaptive Attention-aware Soft Anchor-Point Detector (SASAPD), to identify structures and evaluate their performance. Specifically, in the case of analyzing single-phase metallographic image, MAUNet investigates the difference between low-frequency and high-frequency and prevents duplication of low-resolution information in skip connection used in an U-Net like structure, and incorporates spatial-channel attention module with the decoder to enhance interpretability of features. In the case of analyzing multi-phase metallographic image, SASAPD explores and ranks the importance of anchor points, forming soft-weighted samples in subsequent loss design, and self-adaptively evaluates the contributions of attention-aware pyramid features to assist in detecting elements in different sizes. Extensive experiments on the above two datasets demonstrate the superiority and effectiveness of our two deep neural networks compared to state-of-the-art models on different metrics. Full article

(This article belongs to the Special Issue Deep Learning Image Recognition Systems)

► Show Figures

Figure 1

Journal Menu

Journal Browser

Deep Learning Image Recognition Systems

Share This Special Issue

Special Issue Editors

Special Issue Information

Published Papers (8 papers)

Research

Further Information

Guidelines

MDPI Initiatives

Follow MDPI