arXiv：cs.AI· Vicente Balmaseda, Ching-Long Lin, Tianbao Yang·· 3 小时前
SCION：无需预训练，在单一模型中联合生成与自监督表征学习
From Pixels, Without Pre-training: Joint Generative and Self-Supervised Representation Learning in One Model
AI 导读
SCION 在单一像素空间模型中联合实现生成与自监督表征学习，无需标签与预训练模型。其在 ImageNet 256x256 上以 JiT-B 配方达到 8.92 FID，超过对齐 DINOv2 的 iREPA（46.44）与 RCG（14.27）；JiT-L 下无引导为 5.89 FID，加入表征引导后为 3.47 FID。
正文
Abstract:Strong image generation models are conditioned on class labels, aligned to frozen pretrained encoders, or built on separately trained autoencoders. While effective, generation then depends on supervision or pretraining: labels must be annotated, and encoders or autoencoders pretrained for the target domain. We study joint generative and self-supervised representation learning in a single model, enabling self-conditioned generation without labels or pretrained models. This is challenging because the objectives are mismatched: contrastive learning consumes clean augmented views and favors coarse, invariant semantics, while flow matching consumes noisy images and must preserve the fine detail and spatial layout that contrastive learning discards. We propose SCION (Self-conditioned Generation on Self-supervised representation), whose core is a single pixel-space encoder conditioned on the flow timestep and an embedding. For representation learning, this conditioning embedding is a learned global vector shared across images, with the encoder's [CLS] token yielding the semantic representation trained by the contrastive loss. For generative training, the conditioning embedding is the image's own [CLS] representation, while patch tokens pass through a decoder to predict the image. To sample without a reference image at inference, we jointly learn a prior over the embedding. Gradient-norm balancing and stop-gradient mechanisms enable joint optimization in one run. SCION is self-supervised and self-contained, with no labels or pretrained models. On ImageNet 256x256, with the JiT-B recipe and no representation guidance, SCION reaches 8.92 FID, surpassing class-unconditional iREPA, which aligns to pretrained DINOv2 (46.44), and RCG, which conditions on it (14.27). With JiT-L, SCION achieves 5.89 FID without guidance and 3.47 with representation guidance, outperforming RCG with the ADM recipe (6.24).
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (stat.ML) |
| Cite as: | arXiv:2610.05711 [cs.LG] |
| (or arXiv:2610.05711v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.05711 arXiv-issued DOI via DataCite |
Submission history
From: Vicente Balmaseda [view email]
[v1]
Mon, 5 Oct 2026 02:51:03 UTC (20,250 KB)
来源：arXiv：cs.AI · arxiv.org