Domain-adaptive video classification method and device, equipment, medium and product

The invention relates to a domain-adaptive video classification method and device, equipment, a medium and a product. The method comprises the steps of obtaining video input samples of a source domain and at least one target domain, and performing classification based on features of the video input samples of the source domain to obtain an initial video classification model; constructing at least two private networks, wherein the private networks are used for respectively acquiring semantic irrelevant information features of video input samples in each field; obtaining feature data of video input samples of a source domain and at least one target domain extracted by the initial video classification model, obtaining a feature distribution distance between the video classification model and each private network extraction feature, performing maximization processing on the feature distribution distance, and calculating a maximum mean value difference to obtain a public semantic information feature; and iterative