Cedric Perauer

CedricPerauer's activity

• All
• Models
• Datasets
• Spaces
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes
• Articles

upvoted a paper 1 day ago

Sharp Monocular View Synthesis in Less Than a Second

Paper • 2512.10685 • Published 21 days ago • 22

liked a Space 15 days ago

Running on Zero

Featured

70

SHARP - 3D Gaussian Scene Prediction

🔪

70

Sharp Monocular View Synthesis in Less Than a Second

upvoted a paper 18 days ago

Reflection Removal through Efficient Adaptation of Diffusion Transformers

Paper • 2512.05000 • Published 27 days ago • 15

reacted to toshas's post with 🔥 18 days ago

view post

Post

2199

Introducing 🇨🇭WindowSeat🇨🇭 –– our new method for removing reflections from photos taken through windows, on planes, in malls, offices, and other glass-filled environments.

Finetuning a foundation diffusion transformer for reflection removal quickly runs up against the limits of what existing datasets and techniques can offer. To fill that gap, we generate physically accurate examples in Blender that simulate realistic glass and reflection effects. This data enables strong performance on both established benchmarks and previously unseen images.

To make this practical, the open-source Apache-2 model builds on Qwen-Image-Edit-2509, a 20B image-editing diffusion transformer that runs on a single GPU and can be fine-tuned in about a day. WindowSeat keeps its use of the underlying DiT cleanly separated from the data and training recipe, allowing future advances in base models to be incorporated with minimal friction.

Try it out with your own photos in this interactive demo:
🤗 toshas/windowseat-reflection-removal

Other resources:
🌎 Website: huawei-bayerlab/windowseat-reflection-removal-web
🎓 Paper: Reflection Removal through Efficient Adaptation of Diffusion Transformers (2512.05000)
🤗 Model: huawei-bayerlab/windowseat-reflection-removal-v1-0
🐙 Code: https://github.com/huawei-bayerlab/windowseat-reflection-removal

Team: Daniyar Zakarin ( @daniyarzt )*, Thiemo Wandel ( @thiemo-wandel )*, Anton Obukhov ( @toshas ), Dengxin Dai.
*Work done during internships at HUAWEI Bayer Lab

See translation

🔥

7

7

+

Reply

upvoted an article 28 days ago

view article

Article

Diffusers welcomes FLUX-2

• 
• 
• +6

Nov 25, 2025

•

166

upvoted a paper about 1 month ago

SAM 3: Segment Anything with Concepts

Paper • 2511.16719 • Published Nov 20, 2025 • 123

liked a model about 1 month ago

black-forest-labs/FLUX.2-dev

Image-to-Image • Updated Nov 27, 2025 • 95.7k • • 1.17k

liked a dataset about 1 month ago

p-doom/openai-minecraft-dataset

Updated Aug 5, 2025 • 232 • 6

upvoted a paper about 1 month ago

Kandinsky 5.0: A Family of Foundation Models for Image and Video Generation

Paper • 2511.14993 • Published Nov 19, 2025 • 226

liked a Space about 1 month ago

Running on Zero

Featured

334

Depth Anything 3

🏢

334

Create detailed depth maps from images using Depth Anything 3

upvoted 3 papers about 2 months ago

Time-to-Move: Training-Free Motion Controlled Video Generation via Dual-Clock Denoising

Paper • 2511.08633 • Published Nov 9, 2025 • 54

PAN: A World Model for General, Interactable, and Long-Horizon World Simulation

Paper • 2511.09057 • Published Nov 12, 2025 • 76

Lumine: An Open Recipe for Building Generalist Agents in 3D Open Worlds

Paper • 2511.08892 • Published Nov 12, 2025 • 201

liked a Space about 2 months ago

Running on Zero

443

vggt

🏆

443

VGGT (CVPR 2025)

upvoted 2 papers 2 months ago

Concerto: Joint 2D-3D Self-Supervised Learning Emerges Spatial Representations

Paper • 2510.23607 • Published Oct 27, 2025 • 177

Diffusion Transformers with Representation Autoencoders

Paper • 2510.11690 • Published Oct 13, 2025 • 165

upvoted 4 papers 3 months ago

DiT360: High-Fidelity Panoramic Image Generation via Hybrid Training

Paper • 2510.11712 • Published Oct 13, 2025 • 30

FlashWorld: High-quality 3D Scene Generation within Seconds

Paper • 2510.13678 • Published Oct 15, 2025 • 72

Thinking with Camera: A Unified Multimodal Model for Camera-Centric Understanding and Generation

Paper • 2510.08673 • Published Oct 9, 2025 • 125

Self-Forcing++: Towards Minute-Scale High-Quality Video Generation

Paper • 2510.02283 • Published Oct 2, 2025 • 95