@jo_schb profile picture
jo.schb ✈️CVPR @jo_schb

PhD student @ CompVis Group, LMU Munich

Joined January 2021
  • Tweets 62
  • Following 547
  • Followers 291
26 Photos and videos
Pinned Tweet
jo.schb ✈️CVPR@jo_schb
Apr 27
Diffusion models treat every part of an image equally. → Same number of steps. Same compute. But images aren’t uniform. 🤔 Some regions are easy, others are hard. So why force the model to treat them the same? 🧵
0:05
11
82
588
76,011
I am at @CVPR in Denver this week. If you’re around and want to chat, feel free to send me a DM or stop by one of the two posters I’ll be presenting👇
1
5
445
Denoising, Fast and Slow: Difficulty-Aware Adaptive Sampling for Image Generation 🕐Sun 7th, 3:30 PM - 5:30 PM 📍ExHall A 658 x.com/jo_schb/status/2048765…

jo.schb ✈️CVPR@jo_schb
Apr 27
Diffusion models treat every part of an image equally. → Same number of steps. Same compute. But images aren’t uniform. 🤔 Some regions are easy, others are hard. So why force the model to treat them the same? 🧵
0:05
1
2
235
Probabilistic Precipitation Nowcasting with Rectified Flow Transformers 🕐Sat 6th, 4:45 - 6:45 PM 📍ExHall A & F 401 x.com/jo_schb/status/2062333…

jo.schb ✈️CVPR@jo_schb
Jun 4
⚠️ Standard first stages are not sufficient for safety-critical applications! The most extreme weather events are often the hardest to decode. One latent → many plausible reconstructions Deterministic decoders hide that uncertainty. Meet FREUD 🧵👇
0:02
1
2
214
⚠️ Standard first stages are not sufficient for safety-critical applications! The most extreme weather events are often the hardest to decode. One latent → many plausible reconstructions Deterministic decoders hide that uncertainty. Meet FREUD 🧵👇
0:02
2
7
10
1,271
more replies
For more information: Project page: compvis.github.io/weather-rf Paper: arxiv.org/abs/2605.31204 Code: github.com/CompVis/weather-r…

Probabilistic Precipitation Nowcasting with Rectified Flow Transformers

FREUD enables uncertainty-aware and efficient precipitation nowcasting in latent space.

compvis.github.io
1
3
79
Joint co-first work with @Ragor_ , and collaborators @rmsnorm, Timy Phan. Supervised by Björn Ommer @ CompVis Group LMU Munich
3
90
jo.schb ✈️CVPR retweeted
Stefan Baumann
@StefanABaumann
Jun 1
The internet is full of video. So why can't novel view synthesis just scale on it? Real-world video is simultaneously unposed, messy, and dynamic, breaking self-supervised NVS. We fixed that. RayDer learns static-scene NVS from dynamic internet video, scaling like an LLM. A🧵
0:06
7
35
153
14,181
Diffusion models treat every part of an image equally. → Same number of steps. Same compute. But images aren’t uniform. 🤔 Some regions are easy, others are hard. So why force the model to treat them the same? 🧵
0:05
11
82
588
76,011
more replies
Takeaway 🚀 • Diffusion shouldn’t treat all regions equally • Patch-wise timesteps improve performance, if done right • Allocating compute where it matters gives further gains Project Page: compvis.github.io/patch-forc… arXiv: arxiv.org/abs/2604.19141 Code: github.com/CompVis/patch-for…

1
3
44
2,681
Joint work with @MingGui725184, @Yusong53080064 , @PingchuanMa4, @felix_m_krause, and Björn Ommer! 🫶
6
1,369
Load more