Infinity: Scaling Bitwise AutoRegressive Modeling for High-Resolution Image Synthesis
This addresses the problem of slow and low-quality image generation for AI applications, representing a significant advance rather than an incremental improvement.
The paper tackles high-resolution image synthesis by introducing Infinity, a bitwise autoregressive model that sets new records for text-to-image generation, improving GenEval from 0.62 to 0.73 and ImageReward from 0.87 to 0.96 while generating 1024x1024 images in 0.8 seconds.
We present Infinity, a Bitwise Visual AutoRegressive Modeling capable of generating high-resolution, photorealistic images following language instruction. Infinity redefines visual autoregressive model under a bitwise token prediction framework with an infinite-vocabulary tokenizer & classifier and bitwise self-correction mechanism, remarkably improving the generation capacity and details. By theoretically scaling the tokenizer vocabulary size to infinity and concurrently scaling the transformer size, our method significantly unleashes powerful scaling capabilities compared to vanilla VAR. Infinity sets a new record for autoregressive text-to-image models, outperforming top-tier diffusion models like SD3-Medium and SDXL. Notably, Infinity surpasses SD3-Medium by improving the GenEval benchmark score from 0.62 to 0.73 and the ImageReward benchmark score from 0.87 to 0.96, achieving a win rate of 66%. Without extra optimization, Infinity generates a high-quality 1024x1024 image in 0.8 seconds, making it 2.6x faster than SD3-Medium and establishing it as the fastest text-to-image model. Models and codes will be released to promote further exploration of Infinity for visual generation and unified tokenizer modeling.