A training device, a character recognition device and a method for a character recognition classification network are provided. The device and method train the symmetrical network by constructing sample pairs of unlabeled samples, initialize the classification network by using the parameters of the trained symmetrical network, and train the initialized classification network by using the labeled samples, which can improve the recognition accuracy of the classification network and save the labeling cost effectively.
【技术实现步骤摘要】
【国外来华专利技术】PCT国内申请，说明书已公开。
【技术保护点】
PCT国内申请，权利要求书已公开。
【技术特征摘要】
【国外来华专利技术】PCT国内申请，...
还没有人留言评论。发表了对其他浏览者有用的留言会获得科技券。