Search

Your search keyword '"Liu, Xingmin"' showing total 187 results
Start Over Author "Liu, Xingmin" Language english
187 results on '"Liu, Xingmin"'

39. An Efficient Multi-Label Classification-Based Municipal Waste Image Identification.

Author

Wu, Rongxing, Liu, Xingmin, Zhang, Tiantian, Xia, Jiawei, Li, Jiaqi, Zhu, Mingan, and Gu, Gaoquan

Subjects
TRANSFORMER models, IMAGE recognition (Computer vision), DEEP learning, WASTE management, POWER transformers, ORGANIC wastes, SPINE
Abstract

Sustainable and green waste management has become increasingly crucial due to the rising volume of waste driven by urbanization and population growth. Deep learning models based on image recognition offer potential for advanced waste classification and recycling methods. However, traditional image recognition approaches usually rely on single-label images, neglecting the complexity of real-world waste occurrences. Moreover, there is a scarcity of recognition efforts directed at actual municipal waste data, with most studies confined to laboratory settings. Therefore, we introduce an efficient Query2Label (Q2L) framework, powered by the Vision Transformer (ViT-B/16) as its backbone and complemented by an innovative asymmetric loss function, designed to effectively handle the complexity of multi-label waste image classification. Our experiments on the newly developed municipal waste dataset "Garbage In, Garbage Out", which includes 25,000 street-level images, each potentially containing up to four types of waste, showcase the Q2L framework's exceptional ability to identify waste types with an accuracy exceeding 92.36%. Comprehensive ablation experiments, comparing different backbones, loss functions, and models substantiate the efficacy of our approach. Our model achieves superior performance compared to traditional models, with a mean average precision increase of up to 2.39% when utilizing the asymmetric loss function, and switching to ViT-B/16 backbone improves accuracy by 4.75% over ResNet-101. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

46. Damage evolution monitoring and failure level evaluation of composite bolted joints by MXene/CNTs sensors.

Author

Zhang, Lu, Zhang, Binbin, Lu, Shaowei, Jiang, Xiaowei, Li, Wei, Liu, Xingmin, Ma, Chengkun, and Wang, Xiaoqiang

Subjects
BOLTED joints, CARBON nanotubes, STRUCTURAL failures, COMPOSITE structures, DETECTORS, PROPERTY damage
Abstract

The composite bolted joints are the weak points of overall structure, lacking an effective and accurate monitoring method. In this paper, a novel MXene/CNTs film sensor with high sensitivity and stability was synthesized for the damage evolution monitoring of composite bolted joints. The correlation between the response signals of the sensors and the mechanical behaviors of the composite structures is established. Based on this research, the damage evolution monitoring and failure level evaluation of composite bolted joints were studied. The damage properties of the composite bolted joints can be reflected in real‐time by the resistance change rates of the sensors. The experimental results show that the failure level of composite bolted joints can be divided into four levels (I‐II‐III‐IV) according to the resistance change rates of sensors. Especially when the damage extended to the III stage, the ∆R/R0 surged from 0.3 to 0.75. The time point of this step signal occurred about 1 min earlier than the time point of max bearing strength, which can effectively warn the occurrence of structural failure. Highlights: Research used high‐sensitivity MXene/CNTs sensor to monitor bolt structureSensor array can be integrated with the composite bolted structureStructural damage of composite bolted joints was accurately monitoredFailure level evaluation of composite bolted structure was achieved [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results