直接在像素空间生成图像，避免了传统方法的误差累积问题。
PixNerd: Pixel Neural Field Diffusion
- 用神经场建模图像块解码，实现端到端像素级生成
- 256×256图像上达到2.15 FID，512×512上达2.84 FID
- 适合追求高效高质图像生成的开发者和研究者
当前扩散变压器的成功高度依赖预训练变分自编码器（VAE）构建的压缩潜在空间。然而，这种两阶段训练范式不可避免地引入累积误差和解码伪影。为解决上述问题，研究人员尝试回归像素空间，但代价是复杂的级联流程和更高的标记复杂度。相比之下，我们提出以神经场建模分块解码，构建单一尺度、单阶段、高效且端到端的解决方案——像素神经场扩散（PixelNerd）。得益于PixNerd中高效的神经场表示，我们无需任何复杂级联流程或VAE，直接在ImageNet $256 imes256$ 上取得2.15 FID，$512 imes512$ 上取得2.84 FID。同时，我们将PixNerd扩展至文生图应用，在GenEval基准上取得0.73综合得分，在DPG基准上取得80.9综合得分。
原文摘要 · Abstract (English)
The current success of diffusion transformers heavily depends on the compressed latent space shaped by the pre-trained variational autoencoder(VAE). However, this two-stage training paradigm inevitably introduces accumulated errors and decoding artifacts. To address the aforementioned problems, researchers return to pixel space at the cost of complicated cascade pipelines and increased token complexity. In contrast to their efforts, we propose to model the patch-wise decoding with neural field and present a single-scale, single-stage, efficient, end-to-end solution, coined as pixel neural field diffusion~(PixelNerd). Thanks to the efficient neural field representation in PixNerd, we directly achieved 2.15 FID on ImageNet $256\times256$ and 2.84 FID on ImageNet $512\times512$ without any complex cascade pipeline or VAE. We also extend our PixNerd framework to text-to-image applications. Our PixNerd-XXL/16 achieved a competitive 0.73 overall score on the GenEval benchmark and 80.9 overall score on the DPG benchmark.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。