用变分流匹配实现向量量化图像生成，兼顾连续动态与离散监督。
Purrception: Variational Flow Matching for Vector-Quantized Image Generation
- 在离散码本上学习类别后验，连续空间计算速度场。
- 训练收敛更快，ImageNet-1k上FID达顶尖水平。
- 适合关注生成效率与不确定性建模的研究者。
我们提出Purrception，一种用于向量量化图像生成的变分流匹配方法，能在保持连续传输动态的同时提供显式的类别监督。该方法通过在码本索引上学习类别后验，同时在连续嵌入空间中计算速度场，将连续方法的几何感知能力与离散方法的类别监督优势结合，实现了对合理码本的不确定性量化和温度可控生成。我们在ImageNet-1k 256x256图像生成任务上进行评估，训练收敛速度优于连续与离散流匹配基线，且在FID指标上达到与当前最先进模型相当的性能。结果表明，变分流匹配能有效融合连续传输与离散监督，提升图像生成的训练效率。
原文摘要 · Abstract (English)
We introduce Purrception, a variational flow matching approach for vector-quantized image generation that provides explicit categorical supervision while maintaining continuous transport dynamics. Our method adapts Variational Flow Matching to vector-quantized latents by learning categorical posteriors over codebook indices while computing velocity fields in the continuous embedding space. This combines the geometric awareness of continuous methods with the discrete supervision of categorical approaches, enabling uncertainty quantification over plausible codes and temperature-controlled generation. We evaluate Purrception on ImageNet-1k 256x256 generation. Training converges faster than both continuous flow matching and discrete flow matching baselines while achieving competitive FID scores with state-of-the-art models. This demonstrates that Variational Flow Matching can effectively bridge continuous transport and discrete supervision for improved training efficiency in image generation.
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。