In essence, the artificial intelligence takes a text prompt inputted by the user (in this case, all are titles of novels) and tries to make an image based on it. It is important to note that the interpretations are very, very literal. It is also important to note that in some cases, the training database may have seen images of the book title, so pictures of the cover may appear in some images. All images were generated by VQGAN+CLIP. If you wish to try this out yourself, clear instructions can be found in the source link. Note: Neither the instructions document nor the actual program was created by me.
Slideshow: Type answers based on single images at one time
In order to create a playlist on Sporcle, you need to verify the email address you used during registration. Go to your Sporcle Settings to finish the process.
Comments