Results 151 to 160 of about 4,082 (184)
[Comparison of the efficacy of sigmoid sinus constriction plus cartilage reconstruction and sigmoid sinus return surgery in sigmoid sinus-related pulsatile tinnitus]. [PDF]
Cao W, Wu Z, Zhao Y, Tong B, Yao K.
europepmc +1 more source
[A segmented backprojection tensor degradation feature encoding model for motion artifacts correction in dental cone beam computed tomography]. [PDF]
Zeng Z, Wang Y, Lin Z, Bian Z, Ma J.
europepmc +1 more source
Some of the next articles are maybe not open access.
Related searches:
Related searches:
中国科学与技术学报
本文提出了一种优化后的YOLOv5算法,旨在识别绝缘子缺陷。通过结合Transformer模型、BiFPN(双向特征金字塔网络)和CBAM(卷积块注意力模块)注意力机制,显著提升了识别的准确度、稳定性和泛化能力。此算法能精确识别和定位各种形态和尺寸的缺陷,为绝缘子缺陷检测提供了高效且可靠的解决方案。实验数据显示,优化后的YOLOv5算法在绝缘子故障识别中表现出色,检测准确率高达76.8%,相较于传统YOLOv5算法,性能提升超过20%。在处理复杂绝缘子缺陷和小样本目标识别时,该算法表现出更强的能力 ...
徐铭申 +4 more
semanticscholar +1 more source
本文提出了一种优化后的YOLOv5算法,旨在识别绝缘子缺陷。通过结合Transformer模型、BiFPN(双向特征金字塔网络)和CBAM(卷积块注意力模块)注意力机制,显著提升了识别的准确度、稳定性和泛化能力。此算法能精确识别和定位各种形态和尺寸的缺陷,为绝缘子缺陷检测提供了高效且可靠的解决方案。实验数据显示,优化后的YOLOv5算法在绝缘子故障识别中表现出色,检测准确率高达76.8%,相较于传统YOLOv5算法,性能提升超过20%。在处理复杂绝缘子缺陷和小样本目标识别时,该算法表现出更强的能力 ...
徐铭申 +4 more
semanticscholar +1 more source

