期刊文献+
共找到3篇文章
< 1 >
每页显示 20 50 100
Unifying Convolution and Transformer Decoder for Textile Fiber Identification
1
作者 许罗力 李粉英 常姗 《Journal of Donghua University(English Edition)》 CAS 2023年第4期357-363,共7页
At present,convolutional neural networks(CNNs)and transformers surpass humans in many situations(such as face recognition and object classification),but do not work well in identifying fibers in textile surface images... At present,convolutional neural networks(CNNs)and transformers surpass humans in many situations(such as face recognition and object classification),but do not work well in identifying fibers in textile surface images.Hence,this paper proposes an architecture named FiberCT which takes advantages of the feature extraction capability of CNNs and the long-range modeling capability of transformer decoders to adaptively extract multiple types of fiber features.Firstly,the convolution module extracts fiber features from the input textile surface images.Secondly,these features are sent into the transformer decoder module where label embeddings are compared with the features of each type of fibers through multi-head cross-attention and the desired features are pooled adaptively.Finally,an asymmetric loss further purifies the extracted fiber representations.Experiments show that FiberCT can more effectively extract the representations of various types of fibers and improve fiber identification accuracy than state-of-the-art multi-label classification approaches. 展开更多
关键词 non-destructive textile fiber identification transformer decoder asymmetric loss
下载PDF
基于DCNv2和Transformer Decoder的隧道衬砌裂缝高效检测模型研究
2
作者 孙己龙 刘勇 +4 位作者 周黎伟 路鑫 侯小龙 王亚琼 王志丰 《图学学报》 2024年第5期1050-1061,共12页
为解决因衬砌裂缝性状随机、分布密集、标注框分辨率低所导致的现有模型识别精度低、检测速度慢及参数量庞大等问题,以第2版可变形卷积网络(DCNv2)和端到端变换器解码器(Transformer Decoder)为基础对YOLOv8网络框架进行改进,提出了面... 为解决因衬砌裂缝性状随机、分布密集、标注框分辨率低所导致的现有模型识别精度低、检测速度慢及参数量庞大等问题,以第2版可变形卷积网络(DCNv2)和端到端变换器解码器(Transformer Decoder)为基础对YOLOv8网络框架进行改进,提出了面向衬砌裂缝的检测模型DTD-YOLOv8。首先,通过引入DCNv2对YOLOv8主干卷积网络C2f进行融合以实现模型对裂缝形变特征的准确快速感知,同时采用Transformer Decoder对YOLOv8检测头进行替换以实现端到端框架内完整目标检测流程,从而消除因Anchor-free处理模式所带来的计算消耗。采用自建裂缝数据集对SSD,Faster-RCNN,RT-DETR,YOLOv3,YOLOv5,YOLOv8和DTD-YOLOv8的7种检测模型进行对比验证。结果表明:改进模型F1分数和mAP@50值分别为87.05%和89.58%;其中F1分数相较其他6种模型分别提高了14.16%,7.68%,1.55%,41.36%,8.20%和7.40%;mAP@50分别提高了28.84%,15.47%,1.33%,47.65%,10.14%和10.84%。改进模型参数量仅为RT-DETR的三分之一,检测单张图片的速度为16.01 ms,FPS为65.46帧每秒,对比其他模型检测速度得到提升。该模型在面向运营隧道裂缝检测任务需求时能够表现出高效的性能。 展开更多
关键词 隧道工程 目标检测 第2版可变形卷积网络 transformer Decoder 衬砌裂缝
下载PDF
Optical/digital color photography based on white-light information processing
3
作者 罗罡 刘福来 +4 位作者 林列 方志良 王肇圻 母国光 翁志成 《Science China(Technological Sciences)》 SCIE EI CAS 2001年第2期140-148,共11页
The achievement in optical/digital color photography based on white-light information processing including the color-encoding camera, the color image decoder, the integral window Fourier algorithm of the Fourier trans... The achievement in optical/digital color photography based on white-light information processing including the color-encoding camera, the color image decoder, the integral window Fourier algorithm of the Fourier transform in digital decoding, the color correction of the retrieval color image and the fusion of zero order diffraction is reported. This technique has found its important applications in the fields of aerial reconnaissance photography and far-distance ground photography due to its features of large information capacity, convenience in archival storage, the capability of color enhancement, particularly easy transportation by Internet. 展开更多
关键词 white-light information processing tricolor grating color photography digital Fourier transform decoding
原文传递
上一页 1 下一页 到第
使用帮助 返回顶部