Stephen Jiezhi Yang

@Stepenyang

Research Scientist @ Wayve Labs Prev: Berkeley & Harvard

Sunnyvale, CA

stephenjyang.com

Joined September 2023

84 Following

17 Followers

Replies Reposts Media

Stephen Jiezhi Yang
@Stepenyang
Sep 30
Riding the Wayve!
Alex Kendall
@alexgkendall
Sep 22
Today @wayve_ai and @MercedesBenz are announcing a global strategic partnership! Through this definitive production agreement, we’ll bring our AI Driver to future Mercedes-Benz vehicles starting in the next two years. This is the world's first integration of the Wayve AI Driver
Stephen Jiezhi Yang
@Stepenyang
Sep 30
Mem3R uses fast-weight TTT memory for streaming pose estimation and 3D reconstruction. Great work with the team!
Zhenjun Zhao
@zhenjun_zhao
Apr 10
Mem3R: Streaming 3D Reconstruction with Hybrid Memory via Test-Time Training @liu_changkun, Jiezhi Yang, Zeman Li, Yuan Deng, Jiancong Guo, Luca Ballan tl;dr: implicit fast-weight memory->pose tracking; explicit token-based state->geometric mapping arxiv.org/abs/2604.07279
Stephen Jiezhi Yang
@Stepenyang
Sep 30
Another interesting work from our team exploring autoregressive world modeling in ray space!
Applied Intuition
@AppliedInt
May 29
Current world models often separate spatial reasoning from temporal modeling, leading to consistency bottlenecks and scaling limits. We developed RAYNOVA, a 4D world foundation model, to address this by unifying space and time into a single representation via a pure
Stephen Jiezhi Yang
@Stepenyang
Sep 30
Great work from our team on scalable, label-free pretraining from in-the-wild driving videos— accepted at CVPR this year.
Applied Intuition
@AppliedInt
Sep 24
Training AVs is complex and expensive to scale in new cities and terrains. From calibrated multi-sensor rigs to millions of labeled miles and painstaking annotation, we need to remove the blockers to autonomy development. Meanwhile, there are millions of hours of unlabeled
Stephen Jiezhi Yang
@Stepenyang
Nov 28, 2024
Grateful for the opportunity to contribute to the project and excited to continue pushing the boundaries of 3D vision!
Heng Yang
@hankyang94
Nov 27, 2024
I always thought camera pose estimation is necessary for 3D reconstruction, until Zequn and Stephen proved me wrong! Introducing PreF3R, purely feed-forward 3D Gaussian Splatting without any intermediate pose estimation and COLMAP initialization. Video in, 3D Gaussians and

Stephen Jiezhi Yang

Stephen Jiezhi Yang’s posts