Generative adversarial multi-task multi-element sample automatic labeling method and device

The invention provides a generative adversarial multi-task multi-element sample automatic labeling method, which comprises the steps of S1, guiding a generator to automatically label a to-be-detectedtarget of multiple elements in a sample by adopting a discriminator of a generative adversarial network to generate a target detection label; S2, performing segmentation labeling processing on the target detection label to obtain a target foreground feature and a target background feature; S3, adding mask masks to the target foreground features for processing to obtain instance segmentation marks;and performing same-class label combination on the instance segmentation labels to obtain semantic segmentation labels. The invention further provides a generative adversarial multi-task multi-element sample automatic labeling device, electronic equipment and a computer readable storage medium. 本公开提供了一种生成对抗式多任务多要素样本自动标注方法，包括：S1，采用生成对抗网络的判别器指导生成器对样本中多要素的待测目标进行自动标注，生成目标检测标注；S2，对目标检测标注进行分割标注处理，得到目标前景特征及目标背景特征；S3，将目标前景特征添加mas