Sangwon Jang

@jangsangwon7

PhD student, @kaist_ai

KAIST

agwmon.github.io

Replies Reposts Media

Pinned
Sangwon Jang
@jangsangwon7
Jan 29
What if your video generator could refine itself—at inference time? ❌No new models. ❌No retraining. ❌No external verifier. 💡 Introducing Self-Refining Video Sampling By reinterpreting a pretrained generator (Wan2.2, Cosmos) as a denoising autoencoder, we enable iterative
00:00
3
Sangwon Jang
@jangsangwon7
Jul 6
I'll be presenting Self-Refining Video Sampling with @Jaehyeong_Jo at #ICML2026. Come chat with us about physically plausible video generation! 🗓 Tue, Jul 7 | 2:00 PM–3:45 PM 📍 HALL A #2512 Paper: arxiv.org/abs/2601.18577 Page: agwmon.github.io/self-refine-vi…
Sangwon Jang
@jangsangwon7
Jan 29
What if your video generator could refine itself—at inference time? ❌No new models. ❌No retraining. ❌No external verifier. 💡 Introducing Self-Refining Video Sampling By reinterpreting a pretrained generator (Wan2.2, Cosmos) as a denoising autoencoder, we enable iterative
00:00
1
Sangwon Jang
@jangsangwon7
Jun 11, 2025
🚨Introducing Frame Guidance: Training-Free Guidance for Frame-Level Control in Video Diffusion Models! Most controllable video generation methods need heavy fine-tuning or target specific tasks. ▶️Frame Guidance is training-free and supports diverse applications! 🧵1/8
00:00
2
Sangwon Jang
@jangsangwon7
Mar 14, 2025
🤔Can you really trust what your AI creates? Meet Silent Branding Attack—a sneaky, no-trigger-needed method to secretly poison text-to-image models! 👀 Test yourself: Can you spot which images below have been silently branded?
1