Shuai Shen

A Graph Convolutional Network for Action Recognition in Occluded Skeleton Data

Sicheng Jin, Kai Hu, Shuai Shen, Yongkai Cai et autres

Skeleton-based human action recognition has achieved significant progress, but local occlusions and missing joints in complex environments (e.g., occlusion and low-light conditions) still degrade recognition accuracy and stability. Existing GCN-based methods aggregate features uniformly across joints and lack mechanisms to suppress unreliable …

cn, gb (code pays fourni par la source)

Huawei Cloud Model-as-a-Service on the CloudMatrix384 SuperPod

Ao Xiao, Bangzheng He, Baoquan Zhang, Baoxing Huai et autres

Scaled-out MoE LLMs and scaled-up SuperPods create new systems challenges for production Model-as-a-Service (MaaS), requiring disaggregation, low-latency communication, and decentralized serving. This report presents xDeepServe, the production serving system behind Huawei Cloud's MaaS offering on CloudMatrix384, a 48-server SuperPod with 384 Ascend …

An advanced three stage lightweight model for underwater human detection

Zichen Liao, Kai Hu, Shuai Shen

This study presents StarEye, a lightweight deep learning model designed for underwater human body detection (UHBD) that addresses the challenges of complex underwater environments. The proposed model incorporates several innovative components: a comprehensive underwater dataset construction methodology, a StarBlock-based backbone structure for …

cn, gb (code pays fourni par la source)

A cell screening method for lithium-ion battery grouping based on pre-trained data-driven model with multi-source time series data

Xiang Wang, Jianjun He, Shuai Shen, ZhenJie Liu

cn (code pays fourni par la source)

Action Recognition Based on Multi-Level Topological Channel Attention of Human Skeleton

Kai Hu, Chaowen Shen, Tianyan Wang, Shuai Shen et autres

In action recognition, obtaining skeleton data from human poses is valuable. This process can help eliminate negative effects of environmental noise, including changes in background and lighting conditions. Although GCN can learn unique action features, it fails to fully utilize the prior …

cn (code pays fourni par la source)

An Improved De-noising Algorithm based on Empirical Mode Decomposition for Respiratory Sound Signals

Jinglong Niu, Lin Wang, Shuai Shen, Zhishuai Chen et autres

An enhanced noise-reduction algorithm utilizing empirical mode decomposition (EMD) has been introduced for the analysis of respiratory sound signals. As a method primarily driven by data, EMD offers significant capabilities in filtering out unwanted noise from signals. Combined with the wavelet thresholding …

Feature selection and online discrimination for weak oxygen absorption spectrum

Shuai Shen, Jianjun He, Xiang Wang, En-Ze Hu

cn (code pays fourni par la source)

Relationship between harmonic line shape and temperature and pressure for wavelength modulation spectroscopy

Shuai Shen, Jianjun He, Xiang Wang

The real-time online detection of oxygen concentrations in medicine glass vials has attracted much attention in the field of pharmaceutical manufacturing. Based on wavelength modulation spectroscopy theory, harmonic line shape is an important factor affecting detection accuracy and speed in the process …

cn (code pays fourni par la source)

SGCS: a signal reconstruction method based on Savitzky-Golaysgz filtering and compressed sensing for wavelength modulation spectroscopy

Shuai Shen, Jianjun He

The residual oxygen concentration in pharmaceutical glass vial variously threatens the aseptic properties of encapsulated agents. The demodulated 2nd harmonic signals in the wavelength modulation spectroscopy (WMS) detection system, the data basis of the inversion of oxygen concentration, are inevitably destroyed by …

cn (code pays fourni par la source)

On the effectiveness of the optimally refined proportional sampling testing strategy

F. T. Chan, I.K. Mak, Tsong Yueh Chen, Shuai Shen

Recently, the effectiveness of subdomain testing and random testing has been studied analytically. T.Y. Chen and Y.T. Yu (1994) found that, for the case of disjoint subdomains, as long as the number of test cases selected from each subdomain is proportional to …

hk, au (code pays fourni par la source)