You Should Look at All Objects

Jin, Zhenchao; Yu, Dongdong; Song, Luchuan; Yuan, Zehuan; Yu, Lequan

doi:10.1007/978-3-031-20077-9_20

Cited by 24 publications

(3 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition, [36] extended FPN using the Asymptotic Feature Pyramid Network (AFPN) to enable interactions between non-adjacent layers. Aiming at the limitation of FPN to detect large targets, [17] proposed a refined FPN structure.YOLO-F [7] achieves state-of-the-art performance with single-level features.SFNet [20] combines different levels of features with semantic streams to improve the performance of the FPN model.SAFNet [16] introduces adaptive feature fusion and self-enhancement modules. [5] proposed a parallel FPN structure for bidirectional fusion for target detection.…”

Section: Pyramid Networkmentioning

confidence: 99%

EA-YOLO: Efficient Extraction and Aggregation Mechanismof YOLO for Fire Detection

Wang,

Qian,

et al. 2024

Preprint

View full text Add to dashboard Cite

For fire detection, there are characteristics such as variable sample feature morphology, complex background and dense target, small sample size of dataset and imbalance ofcategories, which lead to the problems of low accuracy and poor real-time performanceof the existing fire detection models. We propose a flame smoke detection model basedon efficient multi-scale feature enhancement, i.e., EA-YOLO. In order to improve theextraction capability of the network model for flame target features, an efficient attentionmechanism is integrated into the backbone network, Multi Channel Attention (MCA), andthe number of parameters of the model is reduced by introducing the RepVB module;at the same time, we design a multi-weighted multidirectional feature neck structure,Multidirectional Feature Pyramid Network (MDFPN), to enhance the model’s flametarget feature information fusion ability; finally, the CIoU loss function is redesigned byintroducing the Slide weighting function to improve the imbalance between difficult andeasy samples. In addition, to address the problem of a small sample size of fire datasets,this paper establishes two fire datasets, Fire-Smoke and Ro-Fire-Smoke, of which thelatter has the model robustness validation function. The experimental results show that themethod of this paper is 6.5% and 7.3% higher compared to the baseline model YOLOv7on the Fire-Smoke and Ro-Fire-Smoke datasets, respectively. The detection speed is 74.6frames per second. It fully demonstrates that the method in this paper has high flamedetection accuracy while considering the real-time nature of the model. The source codeand dataset are located at https://github.com/DIADH/DIADH.YOLO.

show abstract

Section: Pyramid Networkmentioning

confidence: 99%

EA-YOLO: Efficient Extraction and Aggregation Mechanismof YOLO for Fire Detection

Wang,

Qian,

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…However, this simplistic fusion approach can introduce significant information aliasing effects, ultimately impairing the model's object detection capabilities. Consequently, feature interactions may not be adequately nuanced, making it challenging for the model to accurately discern relationships between features at diverse scales, thereby diminishing the model's detection performance [8][9][10]. Furthermore, FPNs often traverse a lengthy path from the top layer to the bottom layer during feature transfer, potentially causing the top layer features to lose critical information as they pass to the lower layers.…”

Section: Introductionmentioning

confidence: 99%

EB-YOLOX:A Balanced One-Stage Object Detection Model Concentrating on Global Features

Wang,

Liu

et al. 2023

Preprint

View full text Add to dashboard Cite

With the increasing complexity of objects and the limitation of hardware resources, it is crucial to design object detection models that are both highly effective and efficient. Although YOLOX, as a leading object detector, strikes a good balance between the number of parameters and performance, it still has some design weaknesses. These include constraints in Bottleneck extracted features, insufficient feature fusion and information loss in FPN (Feature Pyramid Networks), and an uneven trade-off between performance and efficiency in the detection head. These limitations affect the equilibrium between the number of parameters and model performance. To address these issues, this paper presents an object detection model known as EB-YOLOX (Emphasized Balance YOLOX), which incorporates three key elements: Multidirectional Feature Extraction Bottleneck, Decoupled bidirectional Feature Pyramid Network, and Efficient Decoupling head network. With these components, EB-YOLOX effectively tackles model performance challenges and undergoes a comprehensive evaluation on the MS COCO dataset. Experimental results indicate that EB-YOLOX outperforms YOLOX-S by 1.8% in Average Precision (AP) and achieves significant performance improvements across various scenarios, highlighting its excellent balance between the number of parameters and performance.

show abstract

“…The classical FPN network significantly improves the performance of the detection network through the learning of multi-scale features. Subsequent studies [11][12][13][14][15] have used a similar structure. Several studies have shown that low-level features are helpful for identifying larger targets, and the rich texture information contained in low-level features aids in target localization and accurate bounding box generation.…”

Section: Introductionmentioning

confidence: 99%

TIG-DETR: Enhancing Texture Preservation and Information Interaction for Target Detection

Liu,

Wang,

et al. 2023

Applied Sciences

View full text Add to dashboard Cite

FPN (Feature Pyramid Network) and transformer-based target detectors are commonly employed in target detection tasks. However, these approaches suffer from design flaws that restrict their performance. To overcome these limitations, we proposed TIG-DETR (Texturized Instance Guidance DETR), a novel target detection model. TIG-DETR comprises a backbone network, TE-FPN (Texture-Enhanced FPN), and an enhanced DETR detector. TE-FPN addresses the issue of texture information loss in FPN by utilizing a bottom-up architecture, Lightweight Feature-wise Attention, and Feature-wise Attention. These components effectively compensate for texture information loss, mitigate the confounding effect of cross-scale fusion, and enhance the final output features. Additionally, we introduced the Instance Based Advanced Guidance Module in the DETR-based detector to tackle the weak detection of larger objects caused by the limitations of window interactions in Shifted Window-based Self-Attention. By incorporating TE-FPN instead of FPN in Faster RCNN and employing ResNet-50 as the backbone network, we observed an improvement of 1.9 AP in average accuracy. By introducing the Instance-Based Advanced Guidance Module, the average accuracy of the DETR-based target detector has been improved by 0.4 AP. TIG-DETR achieves an impressive average accuracy of 44.1% with ResNet-50 as the backbone network.

show abstract

You Should Look at All Objects

Cited by 24 publications

References 39 publications

EA-YOLO: Efficient Extraction and Aggregation Mechanismof YOLO for Fire Detection

EA-YOLO: Efficient Extraction and Aggregation Mechanismof YOLO for Fire Detection

EB-YOLOX:A Balanced One-Stage Object Detection Model Concentrating on Global Features

TIG-DETR: Enhancing Texture Preservation and Information Interaction for Target Detection

Contact Info

Product

Resources

About