I need to check the user's selfie for the presence of the desired attraction. For each user, the frequency of such a request is from 1 per hour to 1 in half a day (that is, rarely, never streaming). The user uploads a selfie through the Django form and goes to a new level if there is an attraction on the selfie. I heard about the Google Cloud API for this task. Also, video cards are not a problem (yes, Bitcoin), i.e. The option to teach the model itself is suitable if you advise the guides to be sufficiently detailed (have not yet taught machine learning). Moreover, the sights are needed only Kiev at this stage, so that the base is not so huge and have to be used. Well, is there anything else anyone besides Google for this task? I need all this not for work, an enthusiast’s own project. Thank.