Sangwoo Mo
@sangwoomo
Followers
708
Following
2K
Statuses
285
Postdoc @UMich. Past: PhD @kaist_ai, Intern @AIatMeta, @NVIDIAAI. Work on scalable priors for vision, language, and robotics.
Ann Arbor, MI
Joined April 2016
I’m co-organizing the Pixel-level Vision Foundation Models (PixFoundation) Workshop at #CVPR2025. Please join us by submitting papers or attending talks and posters. I’d also love to grab a coffee if you’re interested in the field. Happy New Year, and see you at @CVPR! 😊
Announcing #CVPR2025 Workshop on Pixel-level Vision Foundation Models (PixFoundation)! 🚀 Welcome researchers interested in vision foundation models for pixel-level image/video understanding: segmentation, depth, motion, visual grounding, and more!
0
1
33
RT @hila_chefer: VideoJAM is our new framework for improved motion generation from @AIatMeta We show that video generators struggle with m…
0
192
0
RT @TairanHe99: 🚀 Can we make a humanoid move like Cristiano Ronaldo, LeBron James and Kobe Byrant? YES! 🤖 Introducing ASAP: Aligning Sim…
0
172
0
RT @DJiafei: Can we build a generalist robotic policy that doesn’t just memorize training data and regurgitate it during test time, but ins…
0
67
0
RT @junxian_he: We replicated the DeepSeek-R1-Zero and DeepSeek-R1 training on 7B model with only 8K examples, the results are surprisingly…
0
669
0
@liyzhen2 I think people less care about this since most interesting applications come from conditional generation (e.g., prompt or layout). We may not need complex priors z but simply make c richer when modeling p(x|c,z).
1
0
1
@lucas_prie Amazing work! Does it mean that we should apply ⊥Adam as our default regularizer (esp. in zero-loss regime)?
1
0
0
RT @lucas_prie: I am excited to share our new paper: “Grokking at the Edge of Numerical stability”! We show that floating point errors in…
0
8
0
RT @KyleStachowicz: How can we train high-frequency generalist robot policies with next-token prediction? In new work with @KarlPertsch/@ph…
0
30
0
RT @yen_chen_lin: Video generation models exploded onto the scene in 2024, sparked by the release of Sora from OpenAI. I wrote a blog post…
0
110
0
RT @anand_bhattad: 🧵 1/3 Many at #CVPR2024 & #ECCV2024 asked what would be next in our workshop series. We're excited to announce "How to…
0
24
0
RT @zhou_xian_: Everything you love about generative models — now powered by real physics! Announcing the Genesis project — after a 24-mon…
0
3K
0
RT @younggyoseo: Introducing CoordTok, a scalable video tokenizer that can encode a 128-frame video into only 1k tokens. CoordTok learns a…
0
52
0
RT @GoogleDeepMind: Introducing Genie 2: our AI model that can create an endless variety of playable 3D worlds - all from a single image.…
0
1K
0
RT @CzyangChen: 🎥 Introducing MultiFoley, a video-aware audio generation method with multimodal controls! 🔊 We can ⌨️Make a typewriter sou…
0
43
0
RT @dangengdg: What happens when you train a video generation model to be conditioned on motion? Turns out you can perform "motion prompti…
0
148
0
RT @theworldlabs: We’ve been busy building an AI system to generate 3D worlds from a single image. Check out some early results on our site…
0
727
0
RT @jjpark3D: I’m recruiting PhD students with computer vision, robotics, or ML experience! We especially encourage applicants from physics…
0
88
0
RT @Tim_Dettmers: This is actually a great argument for using MoEs. When I think about MoEs, I think about the cerebellum and its relations…
0
77
0