EURASIP Journal on Image and Video Processing

Table 2 Comparison of models mAP based on different basic CNN

From: Efficient power component identification with long short-term memory and deep neural network

	Faster-RCNN	R-FCN	Proposed model
VGG-19	0.763	0.791	0.813
ResNet-50	0.781	0.797	0.831
ResNet-101	0.789	0.817	0.846
ResNeXt-50	0.791	0.821	0.845
ResNeXt-101	0.821	0.839	0.869

Back to article page