StraIT: Non-autoregressive Generation with Stratified Image Transformer
This addresses the problem of slow image generation for AI and creative applications, offering a faster and higher-quality alternative to existing methods.
The authors tackled high-quality image synthesis by proposing StraIT, a non-autoregressive generative model that outperforms autoregressive and diffusion models, achieving FID scores of 3.96 at 256x256 resolution on ImageNet without guidance and 3.36 with guidance.
We propose Stratified Image Transformer(StraIT), a pure non-autoregressive(NAR) generative model that demonstrates superiority in high-quality image synthesis over existing autoregressive(AR) and diffusion models(DMs). In contrast to the under-exploitation of visual characteristics in existing vision tokenizer, we leverage the hierarchical nature of images to encode visual tokens into stratified levels with emergent properties. Through the proposed image stratification that obtains an interlinked token pair, we alleviate the modeling difficulty and lift the generative power of NAR models. Our experiments demonstrate that StraIT significantly improves NAR generation and out-performs existing DMs and AR methods while being order-of-magnitude faster, achieving FID scores of 3.96 at 256*256 resolution on ImageNet without leveraging any guidance in sampling or auxiliary image classifiers. When equipped with classifier-free guidance, our method achieves an FID of 3.36 and IS of 259.3. In addition, we illustrate the decoupled modeling process of StraIT generation, showing its compelling properties on applications including domain transfer.