CSI-Based Human Activity Recognition With Graph Few-Shot Learning

Zhang, Yong; Chen, Yang; Wang, Yujie; Liu, Qingqing; Cheng, Andong

doi:10.1109/jiot.2021.3103073

Cited by 29 publications

(6 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…VOLUME 11, 2023 OTSU is a method of automatically determining thresholds using the maximum interclass variance, which is a globalbased binary algorithm. When the threshold taken maximizes the variance between classes, the probability of misdivision is the smallest and the division effect is the best [31].…”

Section: A Artificialsar-vessel Datasetmentioning

confidence: 99%

“…In multi-vessel/ship detection and classification, the ability to represent the point of interest more accurately is important, and one way to do it is by utilizing attention. A convolutional block attention module (CBAM) [27] is one of the attention networks that has widely been used to improve the detection capabilities in many applications such as fly species recognition [28], bamboo sticks counting [29], safety helmets wear-ing recognition [30], and human activity recognition [31]. Therefore, by leveraging attention mechanisms such as focusing on essential features and suppressing irrelevant ones, we hope to boost the power of representation.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

CRAS-YOLO: A Novel Multi-Category Vessel Detection and Classification Model Based on YOLOv5s Algorithm

2023

View full text Add to dashboard Cite

Multi-category vessel detection and classification based on satellite imagery attract a lot of attention due to their significant applications in the military and civilian domains. In this study, we generated a new Artificial-SAR-Vessel dataset based on the combination of the FUSAR-Ship dataset and the SimpleCopyPaste method. We further proposed a novel multi-category vessel detection called CRAS-YOLO which consisted of a convolutional block attention module (CBAM), receptive fields block (RFB), and adaptively spatial feature fusion (ASFF) based on YOLOv5s. The proposed CRAS-YOLO improved the feature pyramid network based on the path aggregation network (PANet), which integrates the RFB feature enhancement module and ASFF feature fusion strategy to obtain richer feature information and realize the adaptive fusion of multi-scale features (RA-PANet). At the same time, a CBAM is added to the backbone to accurately locate the vessel location and improve detection capability. The results confirmed that the proposed CRAS-YOLO model reached a precision, recall rate, and mean average precision (mAP) (0.5) of up to 90.4%, 88.6%, and 92.1% respectively. The proposed model also outperformed previous studies' results in another Sar Ship Detection (SSDD) dataset with precision, recall, and mAP scores of up to 97.3%, 95.5%, and 98.7% respectively.

show abstract

Section: A Artificialsar-vessel Datasetmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

CRAS-YOLO: A Novel Multi-Category Vessel Detection and Classification Model Based on YOLOv5s Algorithm

2023

View full text Add to dashboard Cite

show abstract

“…This trend emerges concerning interconnected devices (e.g., worn, implanted, embedded, and swallowed) located in-onand-around the human body forming a network, which is currently being called the Internet of Bodies (IoBs) [26]. This novel field has many applications, such as human activity recognition [27], user authentication [28], and even emotion recognition [29]. This field also encompasses essential studies on the limitations of such sensors, such as time delay and energy consumption issues [30].…”

Section: B Internet Of Bodiesmentioning

confidence: 99%

Bindi: Affective Internet of Things to Combat Gender-Based Violence

Miranda

Rituerto-González

Luis-Mingueza

et al. 2022

IEEE Internet Things J.

View full text Add to dashboard Cite

The main research motivation of this paper is the fight against gender-based violence and achieving gender equality from a technological perspective. The solution proposed in this work goes beyond currently existing panic buttons, needing to be manually operated by the victims under difficult circumstances. Instead, Bindi, our end-to-end autonomous multimodal system, relies on artificial intelligence methods to automatically identify violent situations, based on detecting fear-related emotions, and trigger a protection protocol, if necessary. To this end, Bindi integrates modern state-of-the-art technologies, such as the Internet of Bodies, affective computing, and cyber-physical systems, leveraging i) affective IoT with auditory and physiological commercial off-the-shelf smart sensors embedded in wearable devices, ii) hierarchical multisensorial information fusion, and iii) the edge-fog-cloud IoT architecture. This solution is evaluated using our own dataset named WEMAC, a very recently collected and freely-available collection of data comprising the auditory and physiological responses of 47 women to several emotions elicited by using a virtual reality environment. On this basis, this work provides an analysis of multimodal late fusion strategies to combine the physiological and speech data processing pipelines to identify the best intelligence engine strategy for Bindi. In particular, the best data fusion strategy reports an overall fear classification accuracy of 63.61% for a subject-independent approach. Both a power consumption study and an audio data processing pipeline to detect violent acoustic events complement this analysis. This research is intended as an initial multimodal baseline that facilitates further work with real-life elicited fear in women.

show abstract

“…Baselines. As our method mainly deals with the few-shot learning scenario, we compare our method with recent state-ofthe-art few-shot recognition methods based on CSI, including the CSI-GDAM [44], the ReWiS [45], and the classic prototypical network [40] that is the baseline method. The CSI-GRAM utilizes the graph neural network and attention scheme to enhance few-shot learning, while the ReWiS proposes SVD data processing and applies the prototypical network.…”

Section: A Setupmentioning

confidence: 99%

AutoFi: Towards Automatic WiFi Human Sensing via Geometric Self-Supervised Learning

Yang¹,

Chen²,

Zhao³

et al. 2022

Preprint

View full text Add to dashboard Cite

WiFi sensing technology has shown superiority in smart homes among various sensors for its cost-effective and privacy-preserving merits. It is empowered by Channel State Information (CSI) extracted from WiFi signals and advanced machine learning models to analyze motion patterns in CSI.Many learning-based models have been proposed for kinds of applications, but they severely suffer from environmental dependency. Though domain adaptation methods have been proposed to tackle this issue, it is not practical to collect highquality, well-segmented and balanced CSI samples in a new environment for adaptation algorithms, but randomly-captured CSI samples can be easily collected. In this paper, we firstly explore how to learn a robust model from these low-quality CSI samples, and propose AutoFi, an automatic WiFi sensing model based on a novel geometric self-supervised learning algorithm. The AutoFi fully utilizes unlabeled low-quality CSI samples that are captured randomly, and then transfers the knowledge to specific tasks defined by users, which is the first work to achieve cross-task transfer in WiFi sensing. The AutoFi is implemented on a pair of Atheros WiFi APs for evaluation. The AutoFi transfers knowledge from randomly collected CSI samples into human gait recognition and achieves state-of-the-art performance. Furthermore, we simulate cross-task transfer using public datasets to further demonstrate its capacity for crosstask learning. For the UT-HAR and Widar datasets, the AutoFi achieves satisfactory results on activity recognition and gesture recognition without any prior training. We believe that the AutoFi takes a huge step toward automatic WiFi sensing without any developer engagement while overcoming the cross-site issue.

show abstract

CSI-Based Human Activity Recognition With Graph Few-Shot Learning

Cited by 29 publications

References 29 publications

CRAS-YOLO: A Novel Multi-Category Vessel Detection and Classification Model Based on YOLOv5s Algorithm

CRAS-YOLO: A Novel Multi-Category Vessel Detection and Classification Model Based on YOLOv5s Algorithm

Bindi: Affective Internet of Things to Combat Gender-Based Violence

AutoFi: Towards Automatic WiFi Human Sensing via Geometric Self-Supervised Learning

Contact Info

Product

Resources

About