Learning to Globally Edit Images with Textual Description
This work addresses a novel problem in computational photography for image editing, though it appears incremental as it builds on existing RNN and GAN methods.
The paper tackles the problem of globally editing images using free-form textual instructions by developing three trainable models based on RNN and GAN, with experimental validation on a dataset of around 2000 image pairs collected via Amazon Mechanical Turk.
We show how we can globally edit images using textual instructions: given a source image and a textual instruction for the edit, generate a new image transformed under this instruction. To tackle this novel problem, we develop three different trainable models based on RNN and Generative Adversarial Network (GAN). The models (bucket, filter bank, and end-to-end) differ in how much expert knowledge is encoded, with the most general version being purely end-to-end. To train these systems, we use Amazon Mechanical Turk to collect textual descriptions for around 2000 image pairs sampled from several datasets. Experimental results evaluated on our dataset validate our approaches. In addition, given that the filter bank model is a good compromise between generality and performance, we investigate it further by replacing RNN with Graph RNN, and show that Graph RNN improves performance. To the best of our knowledge, this is the first computational photography work on global image editing that is purely based on free-form textual instructions.