Abel L. Peirson V (Department of Physics) und E. Meltem Tolunay (Department of Electrical Engineering) von der Stanford University stellen mit dem Paper Dank Learning: Generating Memes Using Deep Neural Networks (arxiv.org/abs/1806.04510) “novel meme generation system” vor:
We introduce a novel meme generation system, which given any image can produce a humorous and relevant caption. Furthermore, the system can be conditioned on not only an image but also a user-defined label relating to the meme template, giving a handle to the user on meme content. The system uses a pretrained Inception-v3 network to return an image embedding which is passed to an attention-based deep-layer LSTM model producing the caption – inspired by the widely recognized Show and Tell Model. We implement a modified beam search to encourage diversity in the captions. We evaluate the quality of our model using perplexity and human assessment on both the quality of memes generated and whether they can be differentiated from real ones. Our model produces original memes that cannot on the whole be differentiated from real ones.
Die Ergebnisse gehen (zum Glück) zwar (noch) knapp an der Funktionalität und Pragmatik von Memes vorbei, aber sie zeigen die Möglichkeiten des Deep Learnings an einem nachvollziehbaren Beispiel gut auf. Danke für den Hinweis, Daniel Pfurtscheller (@dpfu):
📯AI can now generate memes 📯
(Spoiler: They are not funny)Abel L Peirson V & E Meltem Tolunay:
Dank Learning. Generating Memes Using Deep Neural Networkshttps://t.co/iNZMuF1qLN pic.twitter.com/EuJowxUjU3— Daniel Pfurtscheller (@dpfu) 13. Juni 2018