Target detection method and device based on multi-task learning model

The invention discloses a target detection method and device based on a multi-task learning model. The method comprises the steps that an image is input into a feature extraction network of the multi-task learning model to extract a multi-layer feature map of the image, the image comprises a target object, the multi-layer feature map comprises a first-layer feature map and a second-layer feature map, and the feature depth of the first-layer feature map is larger than that of the second-layer feature map; inputting the first-layer feature map into a first sub-task network in the multi-task learning model to obtain detection data of a first bounding box of the target object; and inputting the second-layer feature map into a second sub-task network in the multi-task learning model to obtain detection data of a first key point of the target object. 本公开披露了一种基于多任务学习模型的目标检测方法及装置。所述方法包括：将图像输入所述多任务学习模型的特征提取网络，以提取所述图像的多层特征图，所述图像包括目标物，所述多层特征图包括第一层特征图和第二层特征图，所述第一层特征图的特征深度大于所述第二层特征图的特征深度；将所述第一层特征图输入所述多任务学习模型中的第一子任务网络，以获取所