Ceyuan Yang (@CeyuanY) / X

Ceyuan Yang

154 posts

Ceyuan Yang

@CeyuanY

researcher @ bytedance seed

Hong Kong

scholar.google.com/citations?user…

Joined February 2020

Ceyuan Yang
@CeyuanY
Apr 14, 2025
Glad to share Seaweed-7B, a cost-effective foundation model for video generation. Our tech report highlights the key designs that significantly improve compute efficiency and performance given limited resources, achieving comparable quality against other industry-level models. To
00:00
77K
Ceyuan Yang
@CeyuanY
Sep 25, 2023
Personalization means a lot. We thus introduce the lightweight motion LoRA into #AnimateDiff. Fine-grained motions like the following camera control are enabled. Models are available now at GitHub. Just create more motions as you want.
GIF
97K
Ceyuan Yang
@CeyuanY
Dec 15, 2023
Yuwei (@GuoywGuo) just released #AnimateDiff v3 and #SparseCtrl which allows to animate ONE keyframe, generate transition between TWO keyframes and interpolate MULTIPLE sparse keyframes. RGB images and scribbles are supported for now. Github: github.com/guoyww/Animate…
00:00
57K
Ceyuan Yang
@CeyuanY
Nov 10, 2023
Update of #AnimateDiff: now can avoid watermarks and more importantly, support #SDXL i.e., produce high resolution videos (ie 1024x1024x16 frames with various aspect ratios) with/without personalized models (~13GB VRAM for inference). Try it out on github.com/guoyww/Animate….
00:00
53K
Ceyuan Yang
@CeyuanY
Jul 11, 2023
Wanna make your personalized T2I (\eg, LoRA) move? Try our AnimateDiff which animates multiple models once for all without specific tuning. Page: animatediff.github.io
00:00
31K
Ceyuan Yang
@CeyuanY
Nov 29, 2023
Controllability matters for video generation. We thus propose SparseCtrl that enables keyframe (sketch, depth, img) animation/transition/prediction/interpolation by feeding TEMPORALLY SPARSE condition maps. It’s also compatible with #AnimateDiff. Page: guoyww.github.io/projects/Spars…
00:00
40K
Ceyuan Yang
@CeyuanY
Mar 16, 2025
Check out our latest work CameraCtrl II. By carefully collecting and processing data and introducing as little inductive bias as we can, users are allowed to explore the generated world with appealing dynamics and consistency. Together with extension and distillation, CameraCtrl
00:00
15K
Ceyuan Yang
@CeyuanY
May 24, 2024
Control your camera when generating videos here! Training code and inference weights of both #AnimateDiff and #SVD are now available at github.com/hehao13/Camera…
AK
@_akhaliq
Apr 3, 2024
CameraCtrl Enabling Camera Control for Text-to-Video Generation Controllability plays a crucial role in video generation since it allows users to create desired content. However, existing models largely overlooked the precise control of camera pose that serves as a cinematic
00:00
25K
Ceyuan Yang
@CeyuanY
Mar 14, 2025
We propose Long Context Tuning (LCT) for scene-level video generation to bridge the gap between current single-shot generation and real-world narrative video productions. Homepage: guoyww.github.io/projects/long-… Report: arxiv.org/abs/2503.10589
00:00
47K
Ceyuan Yang
@CeyuanY
Apr 3, 2024
Thank @_akhaliq for sharing our work. As an another plug-in module, CameraCtrl could be inserted into our #AnimateDiff to enable camera controlling. Code and models are avaliable at github.com/hehao13/Camera…
AK
@_akhaliq
Apr 3, 2024
CameraCtrl Enabling Camera Control for Text-to-Video Generation Controllability plays a crucial role in video generation since it allows users to create desired content. However, existing models largely overlooked the precise control of camera pose that serves as a cinematic
00:00
19K
Ceyuan Yang
@CeyuanY
Apr 14, 2025
More videos are available at seaweed.video
AK
@_akhaliq
Apr 14, 2025
ByteDance just announced Seaweed-7B on Hugging Face Cost-Effective Training of Video Generation Foundation Model
00:00
15K
Ceyuan Yang
@CeyuanY
Sep 8, 2023
If your GAN related research (e.g., image, video and 3D) is also bottlenecked by its scale, just like me, try our Aurora, a GAN based text-to-image generator. Surely, code and model will be publicly available soon. ArXiv: arxiv.org/pdf/2309.03904…
8.1K
Ceyuan Yang
@CeyuanY
Jan 15, 2025
One-step this time for video generation! check out more demos here seaweed-apt.com
Peter Lin
@peter9863
Jan 15, 2025
Excited to announce our latest research Seaweed-APT. Unlike diffusion and autoregressive models that require iterative sampling, our model generates 1280x720 24fps videos using only a single step (1NFE)! For more information, visit our website: seaweed-apt.com
00:00
4.2K
Ceyuan Yang
@CeyuanY
Nov 19, 2021
Thank @ak92501 for sharing our recent work! Our GenDA could substantially improve the synthesis quality and diversity with few-shot even one-shot target image. Code is coming soon. Stay tuned!
AK
@_akhaliq
Nov 19, 2021
One-Shot Generative Domain Adaptation abs: arxiv.org/abs/2111.09876 project page: genforce.github.io/genda/ method works well even with large domain gaps, and robustly converges within a few minutes for each experiment