摘要 |
Disclosed are a method and a device for image recognition. One of the embodiments of the method comprises: acquiring an image to be recognized containing a target to be recognized; sending the image to be recognized to the server, and receiving identification information and a confidence parameter of a target object returned by the server, the object to be recognized corresponding to the target object, the identification information and the confidence parameter being obtained by means of recognizing the image to be recognized; when the confidence parameter is greater than a confidence threshold, regarding the identification information of the target object as a recognition result; when the confidence parameter is smaller than the confidence threshold, acquiring from a third party platform tagged information associated to the image to be recognized, and regarding the tagged information as a recognition result. This realizes a combination of automatic recognition of the server and the tagged information of a third party, improving the accuracy rate of recognition, and utilizes the tagged information of the third party for training the recognition model corresponding to the machine learning recognition method used by the server, improving the training effect of the recognition model. |