2 citations · 4 across the 26 of their papers we have counts for
21 papers · 1 filter
RAR: Rate-Adaptive Rollout for Time-Consistent Interactive World Models
Yaoyang Liu, Dongyu Yan, Yihua Du +5
Video-based interactive world models aim to enable real-time, user-controllable environments. Yet most existing systems are trained and rolled out at a fixed temporal rate, limitin…
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…
MaineCoon: Pursuing A Real-Time Audio-Visual Social World Model
Lichen Bai, Tianhao Zhang, Shitong Shao +14
As an increasing majority of global video content is consumed on social platforms for interactive social purposes, video generation models built for social worlds are important but…
LIVEditor-14B: Lightning Unified Video Editing via In-Context Sparse Attention
Shitong Shao, Zikai Zhou, Haopeng Li +4
Video editing has evolved toward In-Context Learning (ICL) paradigms, yet the resulting quadratic attention costs create a critical computational bottleneck. In this work, we propo…
MangaFlow: An End-to-End Agentic Framework for Controllable Story to Manga Generation
Muyao Wang, Zeke Xie, Yanhao Chen +2
End-to-end manga generation is a structured visual storytelling task that requires story decomposition, recurring character and scene grounding, page layout design, panel rendering…
Exploring Data-Free LoRA Transferability for Video Diffusion Models
Yuchen Wang, Wenliang Zhong, Lichen Bai +6
Video diffusion models leveraging step distillation or causal distillation have achieved remarkable performance. However, adapting existing LoRAs to these variants remains a critic…
Reflective Flow Sampling Enhancement
Zikai Zhou, Muyao Wang, Shitong Shao +4
The growing demand for text-to-image generation has led to rapid advances in generative modeling. Recently, text-to-image diffusion models trained with flow matching algorithms, su…
Efficient Video Diffusion Models: Advancements and Challenges
Shitong Shao, Lichen Bai, Pengfei Wan +2
Video diffusion models have rapidly become the dominant paradigm for high-fidelity generative video synthesis, but their practical deployment remains constrained by severe inferenc…
CRAFT: Aligning Diffusion Models with Fine-Tuning Is Easier Than You Think
Zening Sun, Zhengpeng Xie, Lichen Bai +3
Aligning Diffusion models has achieved remarkable breakthroughs in generating high-quality, human preference-aligned images. Existing techniques, such as supervised fine-tuning (SF…
FastLightGen: Fast and Light Video Generation with Fewer Steps and Parameters
Shitong Shao, Yufei Gu, Zeke Xie
The recent advent of powerful video generation models, such as Hunyuan, WanX, Veo3, and Kling, has inaugurated a new era in the field. However, the practical deployment of these mo…
Guidance Matters: Rethinking the Evaluation Pitfall for Text-to-Image Generation
Dian Xie, Shitong Shao, Lichen Bai +5
Classifier-free guidance (CFG) has helped diffusion models achieve great conditional generation in various fields. Recently, more diffusion guidance methods have emerged with impro…
PISA: Piecewise Sparse Attention Is Wiser for Efficient Diffusion Transformers
Haopeng Li, Shitong Shao, Wenliang Zhong +4
Diffusion Transformers are fundamental for video and image generation, but their efficiency is bottlenecked by the quadratic complexity of attention. While block sparse attention a…
Optimizing Few-Step Generation with Adaptive Matching Distillation
Lichen Bai, Zikai Zhou, Shitong Shao +5
Distribution Matching Distillation (DMD) is a powerful acceleration paradigm, yet its stability is often compromised in Forbidden Zone, regions where the real teacher provides unre…
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…
A Simple and Efficient Baseline for Zero-Shot Generative Classification
Zipeng Qi, Buhua Liu, Shiyan Zhang +4
Large diffusion models have become mainstream generative models in both academic studies and industrial AIGC applications. Recently, a number of works further explored how to emplo…
Zigzag Diffusion Sampling: Diffusion Models Can Self-Improve via Self-Reflection
Lichen Bai, Shitong Shao, Zikai Zhou +4
Diffusion models, the most popular generative paradigm so far, can inject conditional information into the generation path to guide the latent towards desired directions. However,…
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…
IV-Mixed Sampler: Leveraging Image Diffusion Models for Enhanced Video Synthesis
Shitong Shao, Zikai Zhou, Lichen Bai +2
The multi-step sampling mechanism, a key feature of visual diffusion models, has significant potential to replicate the success of OpenAI's Strawberry in enhancing performance by i…
Not All Noises Are Created Equally:Diffusion Noise Selection and Optimization
Zipeng Qi, Lichen Bai, Haoyi Xiong +1
Diffusion models that can generate high-quality data from randomly sampled Gaussian noises have become the mainstream generative method in both academia and industry. Are randomly…