10 papers
HunyuanImage 3.0 Technical Report
Tencent Hunyuan Foundation Model Team
We present HunyuanImage 3.0, a native multimodal model that unifies multimodal understanding and generation within an autoregressive framework, with its image generation module pub…
ChatUMM: Robust Context Tracking for Conversational Interleaved Generation
Wenxun Dai, Zhiyuan Zhao, Yule Zhong +12
Unified multimodal models (UMMs) have achieved remarkable progress yet remain constrained by a single-turn interaction paradigm, effectively functioning as solvers for independent…
TAGRPO: Boosting GRPO on Image-to-Video Generation with Direct Trajectory Alignment
Jin Wang, Jianxiang Lu, Guangzheng Xu +10
Recent studies have demonstrated the efficacy of integrating Group Relative Policy Optimization (GRPO) into flow matching models, particularly for text-to-image and text-to-video g…
Refinement via Regeneration: Enlarging Modification Space Boosts Image Refinement in Unified Multimodal Models
Jiayi Guo, Linqing Wang, Jiangshan Wang +6
Unified multimodal models (UMMs) integrate visual understanding and generation within a single framework. For text-to-image (T2I) tasks, this unified capability allows UMMs to refi…
SOAR: Self-Correction for Optimal Alignment and Refinement in Diffusion Models
You Qin, Linqing Wang, Hao Fei +4
The post-training pipeline for diffusion models currently has two stages: supervised fine-tuning (SFT) on curated data and reinforcement learning (RL) with reward models. A fundame…
Hierarchical SVG Tokenization: Learning Compact Visual Programs for Scalable Vector Graphics Modeling
Ximing Xing, Ziteng Xue, Zhenxi Li +8
Recent large language models have shifted SVG generation from differentiable rendering optimization to autoregressive program synthesis. However, existing approaches still rely on…