16 papers
SoL-Refiner: Speed-of-Light One-Step Refinement for High-Resolution Video
Haozhe Liu, Tian Ye, Shuchen Xue +9
High-resolution video generation is expensive, as its cost grows rapidly with the number of spatiotemporal tokens. A practical alternative first generates a lower-resolution video…
Sol-Attn: Accelerating Video Generation Inference via On-the-Fly Attention Sparsification
Haopeng Li, Yitong Li, Junsong Chen +8
Diffusion transformers are essential for high-fidelity video generation, but long token sequences make attention a dominant inference bottleneck. Training-free dynamic sparse atten…
SANA-Video 2.0: Hybrid Linear Attention with Attention Residuals for Efficient Video Generation
Junsong Chen, Jincheng Yu, Yitong Li +11
We introduce SANA-Video 2.0, a hybrid video diffusion transformer instantiated at 5B and 14B scales under a unified architecture. Designed to generate high-quality video up to 720p…
Variational Adapter for Cross-modal Similarity Representation
WenZhang Wei, Zhipeng Gui, Dehua Peng +2
The core of vision-language models lies in measuring cross-modal similarity within a unified representation space. However, most image-text matching or multi-class image classifica…
SANA-Streaming: Real-time Streaming Video Editing with Hybrid Diffusion Transformer
Yuyang Zhao, Yicheng Pan, Qiyuan He +6
Real-time streaming video-to-video editing (V2V) is critical for interactive applications such as live broadcasting and gaming, yet it remains a formidable challenge due to the str…
PixelWizard: Towards Efficient High-Fidelity Video Generation at Ultra-Large Spatial Resolution
Wenxue Li, Jingjing Ren, Peng Zhang +6
High-resolution video generation faces a coupled bottleneck of optimization instability and prohibitive computational costs. The massive expansion of the token sequence not only bi…
SANA-WM: Efficient Minute-Scale World Modeling with Hybrid Linear Diffusion Transformer
Haoyi Zhu, Haozhe Liu, Yuyang Zhao +6
We introduce SANA-WM, an efficient 2.6B-parameter open-source world model natively trained for one-minute generation, synthesizing high-fidelity, 720p, minute-scale videos with pre…
Masked Generative Transformer Is What You Need for Image Editing
Wei Chow, Linfeng Li, Xian Sun +14
Diffusion models dominate image editing, yet their global denoising mechanism entangles edited regions with surrounding context, causing modifications to propagate into areas that…
LucidNFT: LR-Anchored Multi-Reward Preference Optimization for Flow-Based Real-World Super-Resolution
Song Fei, Tian Ye, Sixiang Chen +3
Generative real-world image super-resolution (Real-ISR) can synthesize visually convincing details from severely degraded low-resolution (LR) inputs, yet its stochastic sampling ma…
LucidFlux: Caption-Free Photo-Realistic Image Restoration via a Large-Scale Diffusion Transformer
Song Fei, Tian Ye, Lujia Wang +1
Image restoration (IR) aims to recover images degraded by unknown mixtures while preserving semanticsconditions under which discriminative restorers and UNet-based diffusion priors…
Accelerating Diffusion Model Training under Minimal Budgets: A Condensation-Based Perspective
Rui Huang, Shitong Shao, Zikai Zhou +6
Diffusion models have achieved remarkable performance on a wide range of generative tasks, yet training them from scratch is notoriously resource-intensive, typically requiring mil…
MagicDistillation: Weak-to-Strong Video Distillation for Large-Scale Few-Step Synthesis
Shitong Shao, Hongwei Yi, Hanzhong Guo +5
Recently, open-source video diffusion models (VDMs), such as WanX, Magic141 and HunyuanVideo, have been scaled to over 10 billion parameters. These large-scale VDMs have demonstrat…
CoRe^2: Collect, Reflect and Refine to Generate Better and Faster
Shitong Shao, Zikai Zhou, Dian Xie +4
Making text-to-image (T2I) generative model sample both fast and well represents a promising research direction. Previous studies have typically focused on either enhancing the vis…
MagicInfinite: Generating Infinite Talking Videos with Your Words and Voice
Hongwei Yi, Tian Ye, Shitong Shao +10
We present MagicInfinite, a novel diffusion Transformer (DiT) framework that overcomes traditional portrait animation limitations, delivering high-fidelity results across diverse c…
Magic 1-For-1: Generating One Minute Video Clips within One Minute
Hongwei Yi, Shitong Shao, Tian Ye +7
In this technical report, we present Magic 1-For-1 (Magic141), an efficient video generation model with optimized memory consumption and inference latency. The key idea is simple:…
Bag of Design Choices for Inference of High-Resolution Masked Generative Transformer
Shitong Shao, Zikai Zhou, Tian Ye +3
Text-to-image diffusion models (DMs) develop at an unprecedented pace, supported by thorough theoretical exploration and empirical analysis. Unfortunately, the discrepancy between…