Lightnews — Scholar-powered news

Zhenjun Zhao

@ericzzj.bsky.social

1.3K followers 470 following 990 posts

ericzzj1989.github.io PhD from CUHK. 3D vision, SLAM, SfM, Image Matching (https://github.com/ericzzj1989/Awesome-Image-Matching).

github.com

Posts Media Videos Starter Packs

Pinned

Zhenjun Zhao @ericzzj.bsky.social · May 9

🎉 Thrilled to share our CVPR 2025 Award Candidate & Oral paper:

🔹 GlobustVP
Convex Relaxation for Robust Vanishing Point Estimation in Manhattan World

🧱 Global optimality
💥 Tolerates up to 70% outliers
⚡ Fast runtime

📄 Paper: arxiv.org/abs/2505.04788

💻 Code: github.com/WU-CVGL/GlobustVP

1/

1 5 29

Zhenjun Zhao @ericzzj.bsky.social · 9m

tl;dr: point maps from MoGe+dense correspondences from DKM->initial alignment->graph-based refinement with 3D points and normals

Zhenjun Zhao @ericzzj.bsky.social · 9m

MoRe: Monocular Geometry Refinement via Graph Optimization for Cross-View Consistency

Dongki Jung, Jaehoon Choi, Yonghan Lee, Sungmin Eum, Heesung Kwon, Dinesh Manocha

arxiv.org/abs/2510.07119

1 1

Zhenjun Zhao @ericzzj.bsky.social · 10m

UniFField: A Generalizable Unified Neural Feature Field for Visual, Semantic, and Spatial Uncertainties in Any Scene

Christian Maurer, @snehaljauhri.bsky.social, Sophie Lueth, @georgiachal.bsky.social

tl;dr: in title

arxiv.org/abs/2510.06754

Zhenjun Zhao @ericzzj.bsky.social · 1d

Dropping the D: RGB-D SLAM Without the Depth Sensor

Mert Kiray, Alican Karaomer, Benjamin Busam

tl;dr: DAv2+YOLOv11+Key.Net+ORB->static/dynamic processing->ORB-SLAM3

arxiv.org/abs/2510.06216

Zhenjun Zhao @ericzzj.bsky.social · 1d

Visual Odometry with Transformers

@vyuga3d.bsky.social, Duy-Kien Nguyen, Theo Gevers, @cgmsnoek.bsky.social, @martin-r-oswald.bsky.social

tl;dr: DUSt3R encoder->image token embeddings (+camera embeddings)->time/space attention decoder->rotation+translation

arxiv.org/abs/2510.03348

1 8

Zhenjun Zhao @ericzzj.bsky.social · 1d

tl;dr: no explicit 3D representations; masked autoregressive framework; decoder-only rectified flow transformer

Zhenjun Zhao @ericzzj.bsky.social · 1d

Scaling Sequence-to-Sequence Generative Neural Rendering

Shikun Liu, Kam Woh Ng, Wonbong Jang, Jiadong Guo, Junlin Han, Haozhe Liu, Yiannis Douratsos, Juan C. Pérez, Zijian Zhou, Chi Phung, Tao Xiang, Juan-Manuel Pérez-Rúa

arxiv.org/abs/2510.04236

Zhenjun Zhao @ericzzj.bsky.social · 1d

TCB-VIO: Tightly-Coupled Focal-Plane Binary-Enhanced Visual Inertial Odometry

Matthew Lisondra, Junseo Kim, Glenn Takashi Shimoda, Kourosh Zareinia, Sajad Saeedi

tl;dr: focal-plane sensor-processor arrays help MSCKF

arxiv.org/abs/2510.03919

Zhenjun Zhao @ericzzj.bsky.social · 1d

OKVIS2-X: Open Keyframe-based Visual-Inertial SLAM Configurable with Dense Depth or LiDAR, and GNSS

Simon Boche, Jaehyung Jung, Sebastián Barbas Laina, Stefan Leutenegger

tl;dr: multi-sensor OKVIS2+dense volumetric occupancy maps

arxiv.org/abs/2510.04612

Zhenjun Zhao @ericzzj.bsky.social · 1d

Textured Gaussians for Enhanced 3D Scene Appearance Modeling

Brian Chao, Hung-Yu Tseng, Lorenzo Porzi, Chen Gao, Tuotuo Li, Qinbo Li, Ayush Saraf, @jbhuang0604.bsky.social, Johannes Kopf, Gordon Wetzstein, Changil Kim

tl;dr: RGBA texture maps->Gaussians

arxiv.org/abs/2411.18625

1 1

Zhenjun Zhao @ericzzj.bsky.social · 1d

Universal Beta Splatting

Rong Liu, @paigao.bsky.social, Benjamin Planche, Meida Chen, Van Nguyen Nguyen, Meng Zheng, Anwesa Choudhuri, Terrence Chen, Yue Wang, Andrew Feng, Ziyan Wu

tl;dr: radiance field rendering->N-dimensional anisotropic Beta kernels

arxiv.org/abs/2510.03312

Zhenjun Zhao @ericzzj.bsky.social · 3d

FSFSplatter: Build Surface and Novel Views with Sparse-Views within 3min

Yibin Zhao, Yihan Pan, Jun Nan, Jianjun Yi

tl;dr: VGGT for surface reconstruction under sparse view setting

arxiv.org/abs/2510.02691

Zhenjun Zhao @ericzzj.bsky.social · 5d

tl;dr: inconsistent images+blank images->initial videos->video diffusion model (ViewCrafter) guided by semantic embedding->consistent images

Zhenjun Zhao @ericzzj.bsky.social · 5d

UniVerse: Unleashing the Scene Prior of Video Diffusion Models for Robust Radiance Field Reconstruction

Jin Cao, Hongrui Wu, Ziyong Feng, Hujun Bao, Xiaowei Zhou, Sida Peng

arxiv.org/abs/2510.01669

1 1

Zhenjun Zhao @ericzzj.bsky.social · 5d

EC3R-SLAM: Efficient and Consistent Monocular Dense SLAM with Feed-Forward 3D Reconstruction

Lingxiang Hu, Naima Ait Oufroukh, Fabien Bonardi, Raymond Ghandour

tl;dr: XFeat for tracking; VGGT for mapping

arxiv.org/abs/2510.02080

Zhenjun Zhao @ericzzj.bsky.social · 6d

tl;dr: query+posed mapping images->DUSt3R base->query 3D coordinates

Zhenjun Zhao @ericzzj.bsky.social · 6d

A Scene is Worth a Thousand Features: Feed-Forward Camera Localization from a Collection of Image Features

@axelbarroso.bsky.social, Tommaso Cavallari, Victor Adrian Prisacariu, @ericbrachmann.bsky.social

arxiv.org/abs/2510.00978

1 1 2

Zhenjun Zhao @ericzzj.bsky.social · 6d

Semantic Visual Simultaneous Localization and Mapping: A Survey on State of the Art, Challenges, and Future Directions

Thanh Nguyen Canh, Haolan Zhang, Xiem HoangVan, Nak Young Chong

tl;dr: in title

arxiv.org/abs/2510.00783

Zhenjun Zhao @ericzzj.bsky.social · 7d

GaussianLens: Localized High-Resolution Reconstruction via On-Demand Gaussian Densification

Yijia Weng, Zhicheng Wang, @songyoupeng.bsky.social, @saining.bsky.social, Howard Zhou, Leonidas J. Guibas

tl;dr: initial 3DGS+multi-view images->feed-forward densification

arxiv.org/abs/2509.25603

Zhenjun Zhao @ericzzj.bsky.social · 7d

Please also refer to:
bsky.app/profile/eric...

Zhenjun Zhao @ericzzj.bsky.social · Jun 17

Test3R: Learning to Reconstruct 3D at Test Time

Yuheng Yuan, Qiuhong Shen, Shizun Wang, Xingyi Yang, Xinchao Wang

tl;dr: maximize the geometric consistency between the reconstructions generated from multiple image pairs

arxiv.org/abs/2506.13750

Zhenjun Zhao @ericzzj.bsky.social · 7d

tl;dr: state updating->TTT-style online learning; confidence-guided state update->per-token learning rates

1 1

Zhenjun Zhao @ericzzj.bsky.social · 7d

TTT3R: 3D Reconstruction as Test-Time Training

@xingyu-chen.bsky.social, @fanegg.bsky.social, @xiuyuliang.bsky.social, @andreasgeiger.bsky.social, @apchen.bsky.social

arxiv.org/abs/2509.26645

1 8

Zhenjun Zhao @ericzzj.bsky.social · 7d

Benchmarking Egocentric Visual-Inertial SLAM at City Scale

Anusha Krishnan, Shaohui Liu, @pesarlin.bsky.social, Oscar Gentilhomme, David Caruso, Maurizio Monge, Richard Newcombe, Jakob Engel, @marcpollefeys.bsky.social

tl;dr: upgrade LaMAR

arxiv.org/abs/2509.26639

Zhenjun Zhao @ericzzj.bsky.social · 7d

Graphite: A GPU-Accelerated Mixed-Precision Graph Optimization Framework

Shishir Gopinath, Karthik Dantu, Steven Y. Ko

tl;dr: in title

arxiv.org/abs/2509.26581

1 1

Zhenjun Zhao @ericzzj.bsky.social · 8d

tl;dr: Gaussian primitive->continuous field defined on iso-probability surface->submanifold field; variational autoencoder+optimal transport-based Manifold Distance metric