CAP-YOLO: Channel Attention Based Pruning YOLO for Coal Mine Real-Time Intelligent Monitoring

Xu, Zhi; Li, Jingzhao; Meng, Yifan; Zhang, Xiaoming

doi:10.3390/s22124331

Cited by 13 publications

(6 citation statements)

References 46 publications

(71 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In other words, in dark conditions, the weight of infrared features should be greater than that of visible light features. Therefore, how to cleverly fuse features from both modalities is a crucial step in image fusion algorithms [41].…”

Section: Deep Adaptive Fusionmentioning

confidence: 99%

AFIRE: Adaptive FusionNet for Illumination-Robust Feature Extraction in Heterogeneous Imaging Environments

Yu,

Miao,

Sun

et al. 2024

Preprint

View full text Add to dashboard Cite

The fusion of infrared and visible images aims to synthesize a fused image that incorporates richer information by leveraging the distinct characteristics of each modality. However, the disparate quality of input images in terms of infrared and visible light significantly impacts fusion performance. Addressing this issue, we propose a deep adaptive fusion method in this paper, termed Adaptive FusionNet for Illumination-Robust Feature Extraction (AFIRE), which involves interactive processing of two input features and dynamically adjusts fusion weights under varying illumination conditions. Specifically, we introduce a novel interactive extraction structure during the feature extraction stage for both infrared and visible light, enabling the capture of more complementary information. Additionally, we design a Deep Adaptive Fusion module to assess the quality of input features and perform weighted fusion through a channel attention mechanism. Finally, a new loss function is formulated by incorporating the entropy and median of input images to guide the training of the fusion network. Extensive experiments demonstrate that AFIRE outperforms state-of-the-art methods in preserving pixel intensity distribution and texture details. Source code is available in GitHub https://github.com/ISCLab-Bistu/AFIRE.

show abstract

Section: Deep Adaptive Fusionmentioning

confidence: 99%

AFIRE: Adaptive FusionNet for Illumination-Robust Feature Extraction in Heterogeneous Imaging Environments

Yu,

Miao,

Sun

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…To achieve real-time intelligent analysis for coal mine surveillance videos, the Channel-Attention-Based Pruning YOLO and Adaptive Image Enhancement Parameter To achieve real-time intelligent analysis for coal mine surveillance videos, the Channel-Attention-Based Pruning YOLO and Adaptive Image Enhancement Parameter Selection Module were proposed [87]. Another advanced intelligent monitoring system comprises a video acquiring unit, a working face dip angle detection unit, and a coal seam geological detecting instrument [88].…”

Section: Mining Machines Intelligent Monitoringmentioning

confidence: 99%

Development of Surface Mining 4.0 in Terms of Technological Shock in Energy Transition: A Review

Zhironkin

Taran

2023

Energies

View full text Add to dashboard Cite

The expansion of end-to-end Industry 4.0 technologies in various industries has caused a technological shock in the mineral resource sector, wherein itsdigital maturity is lower than in the manufacturing sector. As a result of the shock, the productivity and profitability of raw materials extraction has begun to lag behind the industries of its deep processing, which, in the conditions of volatile raw materials markets, can provoke sectoral crises. The diffusion of Industry 4.0 technologies in the mining sector (Mining 4.0) can prevent a technological shock if they are implemented in all segments, including quarrying (Surface Mining 4.0). The Surface Mining 4.0 technological platform would connect the advanced achievements of the Fourth Industrial Revolution (end-to-end digital artificial intelligence technologies, cyber-physical systems and unmanned production with traditional geotechnology) without canceling them, but instead bringing them to a new level of productivity, resource consumption, and environmental friendliness. In the future, the development of Surface Mining 4.0 will provide a response to the technological shock associated with the acceleration of the digital modernization of the mining sector and the increase in labor productivity, which are reducing the operating costs of raw materials extraction. In this regard, the given review is an attempt to analyze the surface mining digital transformation over the course of the diffusion of Industry 4.0 technologies covered in scientific publications. The authors tried to show the core and frontiers of Surface Mining 4.0 development to determine the production, economic, and social effect of replacing humans with digital and cyber-physical systems in the processes of mineral extraction. Particular attention was paid to the review of research on the role of Surface Mining 4.0 in achieving sustainable development goals.

show abstract

“…The mAP of the model increases by 1.97%, and the recognition performance of the model is significantly improved [33]. Xu et al used a deep channel attention module (DCAM) to enhance YOLOv3's attention to channel information [34]. To reduce the influence of complex background on small target detection, Tan et al embedded an ultra-lightweight subspace attention mechanism (ULSAM) in YOLOv4.…”

Section: Attention Mechanismmentioning

confidence: 99%

Small target detection algorithm based on attention mechanism and data augmentation

Wang

Liu

et al. 2023

Preprint

View full text Add to dashboard Cite

The detection of masks is of great significance to the prevention of occupational diseases such as infectious diseases and dust diseases. For the problems of small target size, large number of targets, and mutual occlusion in mask-wearing detection, a mask-wearing detection algorithm based on improved YOLOv5s is proposed in present work. First, the ultra-lightweight attention mechanism module ECA was introduced in the Neck layer to improve the accuracy of the model. Second, the influence of different loss functions (GIoU, CIoU, and DIoU) on the improved model was explored, and CIoU was determined as the loss function of the improved model. Besides, the improved model adopted the label smoothing method at the data processing stage, which effectively improved the generalization ability of the model and reduced the risk of overfitting. Finally, the influence of data augmentation methods (Mosaic and Mixup) on model performance was discussed, and the optimal weight of data augmentation was determined. The proposed model was tested on the verification set and the mean precision (mAP), precision, and recall are 92.1%, 90.3%, and 87.4%, respectively. The mAP of the improved algorithm is 4.4% higher than that of the original algorithm.

show abstract

CAP-YOLO: Channel Attention Based Pruning YOLO for Coal Mine Real-Time Intelligent Monitoring

Cited by 13 publications

References 46 publications

AFIRE: Adaptive FusionNet for Illumination-Robust Feature Extraction in Heterogeneous Imaging Environments

AFIRE: Adaptive FusionNet for Illumination-Robust Feature Extraction in Heterogeneous Imaging Environments

Development of Surface Mining 4.0 in Terms of Technological Shock in Energy Transition: A Review

Small target detection algorithm based on attention mechanism and data augmentation

Contact Info

Product

Resources

About