Arbitrary-Scale Image Synthesis
This addresses a limitation in image synthesis for applications requiring flexible resolution output, though it is incremental as it builds on existing positional encoding methods.
The paper tackles the problem of generating images at arbitrary scales, including those unseen during training, by proposing scale-consistent positional encodings and inter-scale augmentations, achieving competitive results across a continuum of scales on standard datasets.
Positional encodings have enabled recent works to train a single adversarial network that can generate images of different scales. However, these approaches are either limited to a set of discrete scales or struggle to maintain good perceptual quality at the scales for which the model is not trained explicitly. We propose the design of scale-consistent positional encodings invariant to our generator's layers transformations. This enables the generation of arbitrary-scale images even at scales unseen during training. Moreover, we incorporate novel inter-scale augmentations into our pipeline and partial generation training to facilitate the synthesis of consistent images at arbitrary scales. Lastly, we show competitive results for a continuum of scales on various commonly used datasets for image synthesis.