This AI Can Generate the Other Half of a Picture Using a GPT Model

Author(s): Louis Bouchard

Originally published on Towards AI.

A good AI, like the one used in Gmail, can generate coherent text and finish your phrase. This one uses the same principles in order to complete an image! All done in an unsupervised training with no labels required at all!

OpenAI recently shared a new paper called “Generative Pretraining from Pixels” which is used to predict pixels without incorporating knowledge of the 2-D image structure. They wanted to see if an architecture mainly used for natural language processing could be used with pictures as well to “reconstruct” an image. Just like when Gmail predicts the end of your message, this AI can predict the end of an image!

They used the popular Bidirectional Encoder Representations from Transformers (BERT) technique for Natural Language Processing pre-training developed by Google. Applying the GPT-2 sequence transformer architecture to predict pixels instead of language tokens. These… Read the full blog for free on Medium.

Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming a sponsor.

Published via Towards AI