Chung-Ho Wu

M.S. student @NYCU_official; ex. AI platform intern @Qualcomm. Obsessed with Computer Vision, XR, NeRF, and Robotics.

[Khairallah AL-Awady @eng_khairallah1]

🚨 Anthropic's own team just showed how to actually use Claude Code properly.

25 minutes. free. from the people who built it.

watch the workshop. bookmark it.

you've been using Claude for months without knowing the commands.

Then read the guide below.

[Zhenjun Zhao @zhenjun_zhao]

GaMO: Geometry-aware Multi-view Diffusion Outpainting for Sparse-View 3D Reconstruction

Yi-Chuan Huang, Hao-Jen Chien, @linjohnss, Ying-Huan Chen, @YuLunLiu790407

tl;dr: multi-view outpainting->sparse-view reconstruction

https://t.co/wE3XAMWeun

[Voxel51 @Voxel51]

One of the biggest bottlenecks in deploying visual AI and computer vision is annotation, which can be both costly and time-consuming. Today, we’re introducing Verified Auto Labeling, a new approach to AI-assisted annotation that achieves up to 95% of human-level performance while cutting labeling costs by up to 100,000x and time by 5,000x.

Read the full paper: https://t.co/eKc1sALnV3

[Chung-Ho Wu @kwlilspace]

Nice

[OpenAI Developers @OpenAIDevs]

Image gen is now available in the API!

We’re launching gpt-image-1, making ChatGPT’s powerful image generation capabilities available to developers worldwide starting today.

✅ More accurate, high fidelity images
🎨 Diverse visual styles
✏️ Precise image editing
🌎 Rich world knowledge
🔤 Consistent text rendering

Check out the guide to start building: https://t.co/6XsgOj3oBO

[Chung-Ho Wu @kwlilspace]

Google win this game.

[Sam Altman @sama]

the new version of images in chatgpt is still rolling out, so please try again later today if you dont get a great one :)

[AK @_akhaliq]

ReCamMaster

Camera-Controlled Generative Rendering from A Single Video

[Chung-Ho Wu @kwlilspace]

Really curious about how it works.
Is it tool calling or sth?

[Orange AI @oran_ge]

Google 的新模型可以输出文图混合了，非常离谱
prompt：从0到1帮我做一个奇异博士的原创设计，从概念图的线稿开始，并且每一步都用图片展示

[Orange AI @oran_ge]

Google 的新模型可以输出文图混合了，非常离谱
prompt：从0到1帮我做一个奇异博士的原创设计，从概念图的线稿开始，并且每一步都用图片展示

[AK @_akhaliq]

Diffusion Self-Distillation app is out on Hugging Face

redefines zero-shot customized image generation using FLUX.

DSD is like DreamBooth, but zero-shot/training-free.

It works across any input subject and desired context—character consistency, item/asset adaptation, scene relighting, and more. It even enables the creation of comics/mangas without any effort in fine-tuning or training a personalized model!

[Gordon Wetzstein @GordonWetzstein]

Introducing FLARE @CVPR 2025

FLARE is a feed-forward model that simultaneously estimates high-quality camera poses, 3D geometry, and appearance from sparse uncalibrated images.
1/4

[Simo Ryu @cloneofsimo]

Have you ever thanked ptrblck for his answers on your CUDA problems?

[Chung-Ho Wu @kwlilspace]

See you guys at #Nashville !!!
#CVPR2025

[Min-Hung (Steve) Chen @CMHungSteven]

Thanks @janusch_patas for sharing!
🎉 AuraFusion got accepted to @CVPR 2025! 🎉
Huge congrats to the team!

[🌐 Website] https://t.co/oN3MiZVM9r
[📝 Paper] https://t.co/Zu8RgUd8Un

#CVPR2025 #CVPR #Nvidia #nvidiaresearch #AI #CV #GenerativeAI #3D #diffusion

[Chung-Ho Wu @kwlilspace]

Glad to be tweeted by MrNeRF🥳

[MrNeRF @janusch_patas]

AuraFusion360: Augmented Unseen Region Alignment for Reference-based 360° Unbounded Scene Inpainting

Contributions:
• A novel depth-aware method for generating unseen masks in 360° unbounded 3D scene inpainting, leveraging multi-view information to improve inpainting accuracy.

• An effective integration of accurate reference view unprojection and SDEdit to generate consistent guided RGBs across viewpoints for fine-tuning.

• A comprehensive framework including a new 360° inpainting dataset and capture protocol, enabling high-quality novel view synthesis and quantitative evaluations of inpainted scenes.

[Chung-Ho Wu @kwlilspace]

🤩

[Zhenjun Zhao @zhenjun_zhao]

AuraFusion360: Augmented Unseen Region Alignment for Reference-based 360° Unbounded Scene Inpainting

@kwlilspace, Yang-Jung Chen, Ying-Huan Chen, @jayinnn, Bo-Hsu Ke, Chun-Wei Tuan Mu, Yi-Chuan Huang, Chin-Yang Lin, @CMHungSteven, Yen-Yu Lin, Yu-Lun Liu

[Xingyi Yang @yxy2168]

🚨 Code Drop Alert! 🚨
We've just released the training code for OminiControl! 🎉
Now YOU can build your own OminiControl model, whether it's for 3D, pose-guided, try-on, or any other control task with FLUX model. 🚀
Start creating today!
https://t.co/FaIcZ4LS1g #OminiControl

[Xuan Luo @XuanLuo14]

Thrilled to announce our paper, Quark: Real-time, High-Resolution, and Generalized Neural View Synthesis, has won Best Paper at SIGGRAPH Asia 2024! 🎉

Grateful for our team’s exceptional collaboration. 💪

Explore our work: https://t.co/xDraYkPrBk

#SIGGRAPHAsia2024

[Matthias Niessner @MattNiessner]

Presenting super exciting 3DGS, Avatars & Diffusion papers @SIGGRAPHAsia:

L3DG: Latent 3D Gaussian Diffusion
https://t.co/JMKEh8UVcv

GGHead: Fast and Generalizable 3D Gaussian Heads
https://t.co/NAR7WW0rIc

NPGA: Neural Parametric Gaussian Avatars
https://t.co/jbrCHv7ej1

[Chung-Ho Wu @kwlilspace]

🔥💸

[Zhenjun Zhao @zhenjun_zhao]

FrugalNeRF: Fast Convergence for Few-shot Novel View Synthesis without Learned Priors

Chin-Yang Lin, Chung-Ho Wu, Chang-Han Yeh, Shih-Han Yen, Cheng Sun, Yu-Lun Liu

https://t.co/OohujE6rSw

[Chung-Ho Wu @kwlilspace]

We got rejected with 334 in CVPR2024. Really upset 🥲. Time to move on to ECCV.