user avatar
Angjoo Kanazawa
@akanazawa
Assistant Professor at @Berkeley_EECS, @berkeley_ai. KAIR, @nerfstudioteam. Amazon Scholar @ FAR. Previously advised @WonderDynamics and @LumaLabsAI. she/her.
Berkeley, CA
people.eecs.berkeley.edu/~kanazawa/
Joined June 2011
624
Following
21K
Followers
Posts
  • Pinned
    user avatar
    Angjoo Kanazawa
    @akanazawa
    Apr 21
    Very excited to share this work @davidrmcall did with the fantastic NVIDIA Finland team last year. We have a surprisingly simple, but sample efficient way to post-train a flow model with RL.
    user avatar
    David McAllister
    @davidrmcall
    Apr 17
    We developed a simple, sample-efficient online RL technique for post-training image generation models. We see it as a possible steerable alternative to CFG, driven by any scalar reward, including human preference.
    00:00
    17K
  • user avatar
    Angjoo Kanazawa
    @akanazawa
    Oct 5, 2022
    Releasing 🌟nerfstudio🌟, a plug-and-play python library to easily create your own NeRFs! @nerfstudioteam is a contributor friendly open-source repo with a realtime web viewer that makes it easy to make cool videos 📽️ github.com/nerfstudio-pro… docs.nerf.studio #nerfstudio 1/
    00:00
  • user avatar
    Angjoo Kanazawa
    @akanazawa
    Dec 18, 2020
    View synthesis is super cool! How can we push it further to generate the world *far* beyond the edges of an image? We present Infinite Nature, a method that combines image synthesis and 3D to generate long videos of natural scenes from a single image. infinite-nature.github.io
    00:00
  • user avatar
    Angjoo Kanazawa
    @akanazawa
    Feb 24, 2025
    Exciting news! MegaSAM code is out🔥 & the updated Shape of Motion results with MegaSAM are really impressive! A year ago I didn't think we could make any progress on these videos: shape-of-motion.github.io/results.html Huge congrats to everyone involved and the community 🎉
    00:00
    95K
  • user avatar
    Angjoo Kanazawa
    @akanazawa
    Jun 16, 2023
    Love this problem statement! Reconstruct the world from retinal reflections: world-from-eyes.github.io The original paper from 2004 by Nishino and Nayar was a super inspiring example of creative research for me: cave.cs.columbia.edu/projects/categ… Great to see a 3D version of it!!
    00:00
    151K
  • user avatar
    Angjoo Kanazawa
    @akanazawa
    Feb 28, 2023
    Psyched to finally release our new work that decouples human🧑🏽‍🤝‍🧑🏻 & camera motion🎥 from an unconstrained video✨ We place people in the world, enabling high quality recovery of motion and spatial relationships between people 🤼 CVPR'23: Code & more video: vye16.github.io/slahmr
    00:00
    86K
  • user avatar
    Angjoo Kanazawa
    @akanazawa
    Mar 26, 2021
    New work on real-time rendering of NeRFs. Try it in your browser! alexyu.net/plenoctrees/ We convert NeRF into a PlenOctree, an octree that captures view-dependent effects with spherical basis functions. No MLPs! On desktop it renders at 150fps, 3000x faster than OG NeRF🙂
    00:00
  • user avatar
    Angjoo Kanazawa
    @akanazawa
    May 16, 2019
    Our demo code for CVPR'19 Human Mesh and Motion Recovery (HMMR) is out!!! akanazawa.github.io/human_dynamics/ github.com/akanazawa/huma… Joint work with Jason Zhang, Panna Felsen and Jitendra Malik
    00:00
  • user avatar
    Angjoo Kanazawa
    @akanazawa
    Sep 29, 2022
    Wow this was fast and looks *really* good!!! DreamFusion: Text-to-3D using 2D Diffusion openreview.net/pdf?id=FjNys5c…
    00:00
  • user avatar
    Angjoo Kanazawa
    @akanazawa
    Feb 26, 2020
    Our code for PIFu: Single-View Full Body 3D Digitization, including Clothing from a single input image is now on Github with *MIT* license! Thanks to collaborators S. Saito, Z. Huang, R. Natsume, S. Morishima, and Hao Li from Waseda University and USC shunsukesaito.github.io/PIFu/
    00:00
  • user avatar
    Angjoo Kanazawa
    @akanazawa
    May 14, 2019
    Pixel-aligned Implicit Function (PIFu), a new memory efficient, fully-convolutional 3D representation for recovering a fully textured surface of a clothed person from a single or multi-view image! shunsukesaito.github.io/PIFu/ With Shunsuke S, Zeng H, @r_natsume, Shigeo M, @HaoLi81
    00:00
  • user avatar
    Angjoo Kanazawa
    @akanazawa
    Jul 22, 2020
    New work in #ECCV2020 learning to reconstruct a morphable shape, texture, and viewpoint from an image collection without 3D ground truth *and* 2D keypoints, allowing us to explore new categories like shoes! Congrats to @goelshbhm! arxiv.org/abs/2007.10982 shubham-goel.github.io/ucmr/
    00:00
  • user avatar
    Angjoo Kanazawa
    @akanazawa
    Jun 14, 2019
    Super excited to share that I will be joining UC Berkeley as an assistant professor from fall 2020! I am delighted that I can continue to be part of the wonderful community at Berkeley, and SO grateful for my mentors, friends and family without whom none of this would be possible
  • user avatar
    Angjoo Kanazawa
    @akanazawa
    Aug 12, 2025
    Viser completely changed the way we do research. Before viser, it was hard to visualize 3D/4D data, let alone share it. Now it’s all just in a browser! It’s amazingly powerful and looks awesome. It’s how we render our results and videos. We love it and hope you will too!
    user avatar
    Brent Yi
    @brenthyi
    Jul 31, 2025
    July has been a big month for Viser! - Released v1.0.0😊 - We did some writing Some demos👇
    00:00
    24K