Post-training is the decisive step for converting a pretrained video generator into a production-oriented model that is instruction-following, controllable, and robust over long temporal horizons. This report presents a systematical post-training framework that organizes supervised policy shaping, reward-driven reinforcement learning, and preference-based refinement into a single stability-constrained optimization stack. The framework is designed around practical video-generation constraints, including high rollout cost, temporally compounding failure modes, and feedback that is heterogeneous, uncertain, and often weakly discriminative. By treating optimization as a staged, diagnostic-driven process rather than a collection of isolated tricks, the report summarizes a cohesive recipe for improving perceptual fidelity, temporal coherence, and prompt adherence while preserving the controllability established at initialization. The resulting framework provides a clear blueprint for building scalable post-training pipelines that remain stable, extensible, and effective in real-world deployment settings.
Text-driven person image generation is an emerging and challenging task in cross-modality image generation. Controllable person image generation promotes a wide range of applications such as digital human interaction and virtual try-on. However, previous methods mostly employ single-modality information as the prior condition (e.g. pose-guided person image generation), or utilize the preset words for text-driven human synthesis. Introducing a sentence composed of free words with an editable semantic pose map to describe person appearance is a more user-friendly way. In this paper, we propose H…
CognitionCapturerPro: Towards High-Fidelity Visual Decoding from EEG/MEG via Multi-modal Information and Asymmetric Alignment
Visual Information Processing Laboratory, School of Electronic Engineering, Xidian University（电子科技大学信息处理实验室，电子工程学院）；School of Artificial Intelligence, Beijing Normal University（北京师范大学人工智能学院）
Erase but Preserve: Controllable Removal of Copyrighted Animation Characters via Optimized Semantic Anchors
Institute of Information Engineering, Chinese Academy of Sciences（中国科学院信息工程研究所）；School of Cyber Security, University of Chinese Academy of Sciences（中国科学院大学网络空间安全学院）
College of Artificial Intelligence and Automation, Hohai University（人工智能与自动化学院，河海大学）；College of DaYu, Hohai University（大禹学院，河海大学）；College of Water Conserwancy and Hydropower Engineering, Hohai University（水利水电工程学院，河海大学）；College of Computer Science and Technology, Zhejiang University of Technology（计算机科学与技术学院，浙江工业大学）
Temporally Consistent and Controllable Video Generation of 2D Cine CMR via Latent Space Motion Modeling
Suzhou Institute of Biomedical Engineering and Technology, Chinese Academy of Sciences（苏州生物医学工程与技术研究所，中国科学院）；SyCoIA, IMT Mines Ales（SyCoIA，IMT Mines Ales）
University of California, Irvine（加州大学欧文分校）；Indian Institute of Technology, Kanpur（印度理工学院坎普尔分校）；Google Research（谷歌研究院）
Diffusion probabilistic models have been shown to generate state-of-the-art results on several competitive image synthesis benchmarks but lack a low-dimensional, interpretable latent space, and are slow at generation. On the other hand, standard Variational Autoencoders (VAEs) typically have access to a low-dimensional latent space but exhibit poor sample quality. We present DiffuseVAE, a novel generative framework that integrates VAE within a diffusion model framework, and leverage this to design novel conditional parameterizations for diffusion models. We show that the resulting model equips…
MultiCrafter: High-Fidelity Multi-Subject Generation via Disentangled Attention and Identity-Aware Preference Alignment
College of Computer Science and Technology, Zhejiang University（浙江大学计算机科学与技术学院）；School of Software Technology, Zhejiang University（浙江大学软件技术学院）；Huawei Technologies Ltd（华为技术有限公司）
SenseTime Research（商汤科技研究院）；Centre for Perceptual and Interactive Intelligence Limited（感知与交互智能中心有限公司）
Capitalizing on the recent advances in image generation models, existing controllable face image synthesis methods are able to generate high-fidelity images with some levels of controllability, e.g., controlling the shapes, expressions, textures, and poses of the generated face images. However, these methods focus on 2D image generative models, which are prone to producing inconsistent face images under large expression and pose changes. In this paper, we propose a new NeRF-based conditional 3D face synthesis framework, which enables 3D controllability over the generated face images by imposin…
DanceCrafter: Fine-Grained Text-Driven Controllable Dance Generation via Choreographic Syntax
East China Normal University（东华大学）；Beijing Dance Academy（北京舞蹈学院）；Beijing University of Posts and Telecommunications（北京邮电大学）；New York University Shanghai（纽约大学上海分校）；Zhongguancun Academy（中关村学院）；Zhongguancun Institute of Artificial Intelligence（中关村人工智能研究院）
DOME: Taming Diffusion Model into High-Fidelity Controllable Occupancy World Model
Horizon Robotics；Institute of Software Chinese Academy of Sciences（中国科学院软件研究所）；Institute of Automation Chinese Academy of Sciences（中国科学院自动化研究所）；The University of Hong Kong（香港大学）；University of Pennsylvania（宾夕法尼亚大学）；HKUST(GZ)（香港科技大学（广州））；University of Chinese Academy of Sciences（中国科学院大学）；HKUST（香港科技大学）
Center For Automation Research, University of Maryland, College Park（马里兰大学帕克分校自动化研究中心）；Microsoft Research（微软研究院）
We present FusedGAN, a deep network for conditional image synthesis with controllable sampling of diverse images. Fidelity, diversity and controllable sampling are the main quality measures of a good image generation model. Most existing models are insufficient in all three aspects. The FusedGAN can perform controllable sampling of diverse images with very high fidelity. We argue that controllability can be achieved by disentangling the generation process into various stages. In contrast to stacked GANs, where multiple stages of GANs are trained separately with full supervision of labeled inte…
AlignVTOFF: Texture-Spatial Feature Alignment for High-Fidelity Virtual Try-Off
Sino-European School of Technology, Shanghai University（上海大学 sino-欧洲技术学院）；National Demonstration Center for Experimental Engineering Training Education, Shanghai University（上海大学国家级实验工程培训教育示范中心）
Diff-PC: Identity-preserving and 3D-aware Controllable Diffusion for Zero-shot Portrait Customization
School of Electronic Science and Engineering, Nanjing University（南京大学电子科学与工程学院）；School of Business, Hohai University（河海大学商学院）；School of Artificial Intelligence, Nanjing University of Information Science and Technology（南京信息工程大学人工智能学院）
DriveGEN: Generalized and Robust 3D Detection in Driving via Controllable Text-to-Image Diffusion Generation
FNii-Shenzhen（深圳未来网络研究院）；SSE, CUHK-Shenzhen（香港中文大学（深圳）理工学院）；National University of Singapore（新加坡国立大学）；Nanyang Technological University（南洋理工大学）；Baoji University of Arts and Sciences（宝鸡文理学院）；Sun Yat-sen University（中山大学）
Tuning-Free Image Editing with Fidelity and Editability via Unified Latent Diffusion Model
Communication University of China（中国传媒大学）；National University of Singapore（新加坡国立大学）；University of California at Merced（加州大学默塞德分校）；Yonsei University（延世大学）
Controllable Skin Synthesis via Lesion-Focused Vector Autoregression Model
School of Translational Medicine, Faculty of Medicine, Nursing and Health Sciences, Monash University（转化医学学院，医学、护理与健康科学学院，墨尔本大学）；Melbourne Sexual Health Centre, Alfred Health（墨尔本性健康中心，阿尔弗雷德健康中心）；AIM for Health Lab, Monash University（健康促进实验室，墨尔本大学）；Faculty of IT, Monash University（信息技术学院，墨尔本大学）；Faculty of Engineering, Monash University（工程学院，墨尔本大学）；Faculty of Infectious and Tropical Diseases, London School of Hygiene and Tropical Medicine（传染病与热带医学学院，伦敦热带医学学院）
本文提出 LF-VAR 模型，通过病灶测量评分与类型标签引导，结合多尺度 VQVAE 和 VAR Transformer，实现基于语言提示的可控高保真皮肤图像合成，FID 平均达 0.74，较 SOTA 提升 6.3%。
