To solve the problems of insufficient sample quantity and quality in military target image recognition training,the CycleGAN model is improved and an effective augmented method is proposed.According to the characteristics of military target samples,the generator structure and loss function structure of the model are modified to improve the stability of the model and the quality of generated images.The image recognition model is trained with the expanded sample data sets,it is found that the images generated by the proposed model can effectively improve the accuracy of the recognition model,the practicability and feasibility of the proposed method in expanding military target samples is proved.
关键词
样本增广/CycleGAN/军事目标/图像生成/图像分类
Key words
data augmentation/CycleGAN/military objective/image generation/image classification
维普
万方数据