Daniel Zoran (@DanielZoran_) header
Daniel Zoran (@DanielZoran_) profile picture

Research Scientist at @GoogleDeepMind my personal account is @zdanielz if you like baked goods and can optionally read Hebrew.

London, England
Joined July 2021
Photo by Daniel Zoran (@DanielZoran_)Photo by Daniel Zoran (@DanielZoran_)Photo by Daniel Zoran (@DanielZoran_)Photo by Daniel Zoran (@DanielZoran_)Photo by Daniel Zoran (@DanielZoran_)
Come see our poster at #CVPR2026 today! Morning session 11:45am, poster 326!
Popular self-supervised models like V-JEPA v2 avoid pixel reconstruction, claiming it's "irrelevant" for high-level semantics. But what if pixels were never a problem? Introducing Recurrent Video Masked Autoencoders (RVM) to be presented at #CVPR2026 arxiv.org/abs/2512.13684 🧵
2
6
782
Looped transformers are all the rage, but the value of recurrence is fundamental, not just hype- after all, the brain is highly recurrent. In that spirit, I'm thrilled to share RVM (to be presented at #CVPR2026): a new class of SoTA recurrent self-supervised video encoders! 🧵
Popular self-supervised models like V-JEPA v2 avoid pixel reconstruction, claiming it's "irrelevant" for high-level semantics. But what if pixels were never a problem? Introducing Recurrent Video Masked Autoencoders (RVM) to be presented at #CVPR2026 arxiv.org/abs/2512.13684 🧵
2
1
1
35
9,267
Popular self-supervised models like V-JEPA v2 avoid pixel reconstruction, claiming it's "irrelevant" for high-level semantics. But what if pixels were never a problem? Introducing Recurrent Video Masked Autoencoders (RVM) to be presented at #CVPR2026 arxiv.org/abs/2512.13684 🧵
2
29
3
195
26,409
The results are promising: ✅ Competitive with V-JEPA v2 on spatio-temporal tasks. ✅ Competitive with DINOv2 on spatial tasks. ✅ Strong performance in the small-model regime WITHOUT distillation.
1
1
8
1,723
Excited to announce MooG for learning video representations. MooG allows tokens to move “off-the-grid” enabling better representation of scene elements, even as they move across the image plane through time. 📜arxiv.org/abs/2411.05927 🌐moog-paper.github.io/
1
15
41
3,953
That was easily one of the best talks I've been to at a conference, fantastic work! at @CVPR
1
2
13
1,034
Today @ICLR_conf: Perceiver IO, an any modal attention architecture that scales gracefully to large inputs and outputs, allowing it to produce excellent results on tasks in vision, language, multimodal domains, and beyond: dpmd.ai/ICLR-Perceiver #ICLR2022 1/
It’s that time of year again…@ICLR_conf! Here’s a brief glimpse into our upcoming oral, spotlight, and poster presentations this week: dpmd.ai/ICLR2022 Don’t miss our 5+ sessions kicking off this morning at 10:30AM BST. See you there! #ICLR2022
1
5
1
31
What's this ticking sound? Ah, it's everyone refreshing the OpenReview website.
1
3