新方法增强用于图像生成的视觉自回归模型

报道来源 [4]

1. arXiv cs.LG TIER_1 English(EN) · Jaehyeon Moon, Bumsub Ham · 2026-06-16 04:00

   面向视觉自回归模型的移位求和量化

   arXiv:2606.16131v1 Announce Type: cross Abstract: Post-training quantization (PTQ) enables efficient deployment of deep networks using a small set of data. Its application to visual autoregressive models (VAR), however, remains relatively unexplored. We identify two key challenge…

2. Hugging Face Daily Papers TIER_1 English(EN) · 2026-06-16 00:00

   统一的自回归多模态建模与共享的上下文-视觉标记器是实现统一的关键

   UniAR presents a unified autoregressive framework that uses a single discrete visual tokenizer to bridge visual understanding and generation, achieving state-of-the-art results in image generation and editing through multi-level feature fusion, bitwise quantization, and parallel …

3. arXiv cs.CV TIER_1 English(EN) · Wujian Peng, Lingchen Meng, Yuxuan Cai, Xianwei Zhuang, Yuhuan Yang, Rongyao Fang, Chenfei Wu, Junyang Lin, Zuxuan Wu, Shuai Bai · 2026-06-17 04:00

   共享上下文视觉标记器的统一多模态自回归建模是实现统一的关键

   arXiv:2606.18249v1 Announce Type: new Abstract: Unified Multimodal Modeling aims to integrate visual understanding and generation within a single system. However, existing approaches typically rely on two disparate visual tokenizers, which splits the representation space and hind…

4. arXiv cs.CV TIER_1 English(EN) · Shuai Bai · 2026-06-16 17:59

   共享上下文视觉标记器的统一多模态自回归建模是实现统一的关键

   Unified Multimodal Modeling aims to integrate visual understanding and generation within a single system. However, existing approaches typically rely on two disparate visual tokenizers, which splits the representation space and hinders truly unified modeling. We propose UniAR, a …