iiinova — where your digital self keeps evolving

PCAsplat: Gaussian Splatting with Local PCA Regularization Vitor Matias, Filipe Nascimento, Kiyohiro Nakayama, João Paulo Lima, Márcus Lobo, Gordon Wetzstein, Leonidas Guibas, Afonso Paiva, Tiago Novello arxiv.org/abs/2610.11011

100

Fast Pose Tracking of Rigid Objects with Compact Pose Graph Optimization Xiaojie Zhang, Tom Fischer, Viktor Larsson, Eddy Ilg tl;dr: RoMa v2->RANSAC Umeyama->pose with uncertainty->uncertainty-aware graph->iSAM2 arxiv.org/abs/2610.11815

Slot3R: Set-Associative Spatial Memory for Streaming 3D Reconstruction Xiyuan Zhang, Yanming Yang, Kaiyuan Xu, Ruibo Li, Chi Zhang tl;dr: preserve distinct local states while selectively consolidating consistent observations arxiv.org/abs/2610.12282

000

Beyond Spatio-Temporal Priors: A Generalizable Approach for Dense Correspondence Matching Luping Liu, Bingyi Kang, Yifan Wang, Dong Xu tl;dr: FLUX2-4B+DINOv3+supervised learning on classical, tracked video, synthetic datasets+teacher-guided iterative refinement arxiv.org/abs/2610.12421

000

S2Tok: Streaming 3D Gaussian Reconstruction with Persistent Spatial Tokens Fang Li, Jiraphon Yenphraphai, Quentin Herau, Depu Meng, Yihan Hu, Tianshuo Xu, Narendra Ahuja, Wei Zhan arxiv.org/abs/2610.08978

100

DynStream: Online Streaming 4D Gaussian Reconstruction of Dynamic Worlds from Unposed Video Dingwei Xian, Xiaoyu Zhou, Yajiao Xiong, Yongtao Wang, Ming-Hsuan Yang tl;dr: incremental integration of incoming observations + time-varying Gaussians arxiv.org/abs/2610.09720

000

DeltaSplat: Iterative Gaussian Refinement for Pose-Free Feed-Forward 3D Gaussian Splatting Chanung Park, Seunghyeon Song, Joo Chan Lee, Eunbyung Park, Jong Hwan Ko tl;dr: condition image-space residuals on per-pixel Plucker rays and rendered depth arxiv.org/abs/2610.09853

Video Prediction Policy 2: Predict Better, Act Better tl;dr: WAM->zero-shot generalization in video prediction and action generation arxiv.org/abs/2610.10270

100

Less Context, Better Geometry: Masked Geometric Encoder for Robust 3D Foundation Models Zhimin Shao, Xijun Liu, Zhaoliang Zhang, Yutao Tang, Abhay Yadav, Rama Chellappa, Cheng Peng arxiv.org/abs/2610.06813

110

DensiTok: Making Feed-Forward 3D Gaussian Splatting See More Views Than It Is Given Minhyeok Lee, Jungho Lee, Minseok Kang, Heeseung Choi, Ig-Jae Kim, Sangyoun Lee tl;dr: complete geometry tokens of unobserved views using rectified flow matching arxiv.org/abs/2610.07958

000

Image-Space Refraction Correction for Underwater 3D Reconstruction: Warping Flat-Port Views into Pinhole Perspective Chelim Lim, @tobiasrobotics.bsky.social, Emilio Olivastri, Beverley Gorry, Michael Milford, Alejandro Fontan arxiv.org/abs/2610.07788

100

MoSE3: Learning World-Space SE(3) at Every Pixel Jiahuan Cheng, Zhiyi Li, Tian Xia, @ruojin.bsky.social, @yilundu.bsky.social, @qianqianwang.bsky.social tl;dr: π3+tracking branch->3D point tracks+rigidity embeddings; differentiable Horn fit->recover per-pixel SE(3) arxiv.org/abs/2610.03716

000

Hierarchy-GBP: Accelerating Factor Graph Inference via Abstraction and Recovery Yuzhou Cheng, Tom Yates, Ignacio Alzugaray, Danyal Akarca, Pedro A. M. Mediano, @ajdavison.bsky.social tl;dr: local message passing + a low-rank coarse correction arxiv.org/abs/2610.06978

000

EvoMem-VLA: State-Evolution Memory for Long-Horizon Robot Manipulation Yuheng Na, Zhide Zhong, Junjie He, Junfeng Li, Haodong Yan, Jiaan Wang, Jiaguan Zhu, Yangyang Zheng, Tianyu Huang, Haoang Li tl;dr: encode and retain observed changes between historical states arxiv.org/abs/2610.05418

EmbodiedSmith: Scaling Embodied Data through Recursive Self-Improvement Flywheel in Simulation tl;dr: in title arxiv.org/abs/2610.07969

100

LoCoSplat: Real-Time Feed-Forward 3D Gaussian Splatting with Minimal 3D Reasoning Sinan Wang, Jinjin He, Yuchen Sun, Duowen Chen, Shenyifan Lu, Bo Zhu tl;dr: 3D stage has no learned 3D network arxiv.org/abs/2610.04351

000

VASC: Value-Aware Sparse Attention with Cross-Layer Memory for Efficient 3D Reconstruction Junyi Wu, Fanqing Kong, Leyang Chen, Shaoqiu Zhang, Yulun Zhang tl;dr: value contrast guides block selection; cross-layer memory tracks unserved demand arxiv.org/abs/2610.01013

000

ARROW: Arbitrary Reconstruction and Tracking of 4D Observations in the Wild @ifradlin.bsky.social, Christian Schmidt, Jens Piekenbrinck, @karimknaebel.bsky.social, Gonzalo Martin Garcia, Bastian Leibe arxiv.org/abs/2610.01314

100

EvenSplat: Coupled 2D-3D Decomposition for Gaussian Splatting under Exposure and Illumination Variation Tongyu Wu, Jacob Edwards, Ziteng Cui, Caigui Jiang, Cheng Wang tl;dr: image-space decomposition optimized together with the Gaussian representation arxiv.org/abs/2610.01876

CLoSeR: Closing the Loop for Long-Context Streaming Reconstruction Moyang Li, Zihan Zhu, Wei Zhang, @marcpollefeys.bsky.social, @danielbarath.bsky.social tl;dr: LoGeR + loop closure arxiv.org/abs/2610.01927

011

UniWAM: Unified World-Action Model Jiayi Chen, Wenxuan Song, Jingbo Wang, Shuai Zhou, Xicheng Gong, Zehua Fan, Ziyang Zhou, Junwu E, Haodong Yan, Fuhao Li, Qize Yu, Xu Huang, Pengwei Wang, Wen Chen, Shunbo Zhou, Haoang Li arxiv.org/abs/2610.02054

100

Reconstructing the Dynamic World: A Representation-Centric View of 4D Scene Reconstruction tl;dr: survey arxiv.org/abs/2609.39960

100

FAST-Sync: Fast Group Synchronization for any Matrix Lie Group Shane Holmes, Yiran Luo, Firat Taxpulat, David M. Rosen, @fdellaert.bsky.social arxiv.org/abs/2609.38594

100

LBDU-VIO: Learned Bias Dynamics and Uncertainty for Visual-Inertial Odometry with Unreliable Vision Qizhi Guo, Junning Lyu, Defu Lin, Shaoming He tl;dr: ODE model->continuous time bias dynamics; IMU uncertainty model->measurement noise covariance arxiv.org/abs/2609.39125

000

Efficient Multi-Modal Planning with Reward-Guided Preference Optimization for Autonomous Driving Chenglin Chen, Lujia Wang, Xinhu Zheng, Jun Ma, Haoang Li tl;dr: sparse anchors + an offset module to refine trajectories arxiv.org/abs/2609.38862

000

CSWAM: Better Causal Semantic Representations for Out-of-Distribution Generalization in World Action Models Tianbin Liu, Jian Zhu, Taiyi Su, Jianjun Zhang, Chong Ma, Zitai Huang, Weiyi Lu, Yi Xu arxiv.org/abs/2609.18462

100

Zhenjun Zhao @ericzzj.bsky.social · 30/09/2026

Discrete Forcing: Infusing Discrete Guidance into Continuous Denoising for Few-Step Action Experts Jingbo Wang, Wenxuan Song, Wenhao Yu, Han Zhao, Xi Wang, Jiayi Chen, Donglin Wang, Yan Wang, Haoang Li tl;dr: in title arxiv.org/abs/2609.39526

000

A QCQP-Representable IMU Pre-Integration Factor for Certifiable State Estimation Utkarsh Rai, Zhexin Xu, Bang-Shien Chen, David Rosen tl;dr: Cayley map->IMU pre-integration factor in quadratic formulation; redundant constraint in rotation/wedge/translation arxiv.org/abs/2609.38048

000

Zhenjun Zhao @ericzzj.bsky.social · 30/09/2026

Pow3R-SLAM: Real-Time RGB-D SLAM with 3D Reconstruction Priors Christopher Kolios, Ishaan Mehta, Sasa Janjic, Yeganeh Bahoo, Sajad Saeedi tl;dr: MASt3R-SLAM + depth prior arxiv.org/abs/2609.38054

000

Zhenjun Zhao @ericzzj.bsky.social · 30/09/2026

MVG-WAM: Multiple View Geometry-Aware World-Action Modeling for Robotic Manipulation Wenbo Chen, Tianfu Li, Haoxuan Xu, Zhihao Cao, Zhenghan Chen, Zhengming Zhu, Zizhou Luo, Guosheng Yang, Yuan Liu, Lujia Wang, Wen Chen, Haoang Li arxiv.org/abs/2609.37793

100

QuacamFM: Quaternion-Constrained Flow Matching for Camera Pose Estimation Bao-Long Tran, Cuong Le, Tahereh Dehdarirad, Fredrik Viksten, Per-Erik Forssén arxiv.org/abs/2609.32455

120

From Feed-Forward to Flow: Unifying Reconstruction and Generation Is Easier Than You Think Haoru Wang, Qianfan Shen, Kai Ye, Wenzheng Chen, Baoquan Chen tl;dr: shared clean-target predictor->feed-forward reconstruction and conditional generation arxiv.org/abs/2609.32761

Many Eyes, One World: Feed-Forward 3D Reconstruction from Mixed Cameras Qiaoge Li, Yifan Zhan, Haijun Yang, Haiyang Liu, Yiyi Cai, Chenchi Luo tl;dr: perspective, fisheye and full-panorama images in a single forward pass arxiv.org/abs/2609.35658

021

Zhenjun Zhao @ericzzj.bsky.social · 28/09/2026

SLIP-VLA: Single-Step Latent Imagination for Policy Learning in Vision-Language-Action Models Tianfu Li, Haoxuan Xu, Wenbo Chen, Haitian Li, Changchuan Yang, Xinhu Zheng, Jun Ma, Yuan Liu, Lujia Wang, Haoang Li arxiv.org/abs/2609.33575

100

Light Field Primitive for Novel View Synthesis Liang Chen, Jiahui Ning, Xun Jiang, Xing Xu, Jimmy Ren, Fenglei Fan, Heng Tao Shen tl;dr: scene->primitives defined in two-plane parameterization arxiv.org/abs/2609.31198

000

Zhenjun Zhao @ericzzj.bsky.social · 28/09/2026

VkVIO: Cross-platform GPU Acceleration for Visual-Inertial Odometry with Vulkan Ole Hoffmann, Mateo de Mayo, @dcremers.bsky.social tl;dr: Basalt->Vulkan arxiv.org/abs/2609.30459

000

Zhenjun Zhao @ericzzj.bsky.social · 28/09/2026

TrackEverything: Long Horizon Dense Tracking via De-Duplicating 3D Scene Representations Ayush Jain, Sreeharsha Paruchuri, Ishita Gupta, Fan Zhang, Tanner Schmidt, Jakob Engel, Katerina Fragkiadaki, Adam W. Harley arxiv.org/abs/2609.30222

100

Zhenjun Zhao @ericzzj.bsky.social · 24/09/2026

DAVIO: Dense Monocular-Inertial SLAM with Feed-Forward Initialization and Pose-Conditioned Mapping Jaafar Mahmoud, Arthur Movsesyan, Mikhail Iumanov, Sergey Kolyubin tl;dr: DAv3 + OpenVINS arxiv.org/abs/2609.27702

000

Zhenjun Zhao @ericzzj.bsky.social · 24/09/2026

Kairos: Grounded Forecasting of Presence and Directional Flow in 4D Scene Graphs Iacopo Catalano, Julio A. Placed, @jcivera.bsky.social, Jorge Peña Queralta tl;dr: 3DSG->predictive directional-flow memory->4DSG arxiv.org/abs/2609.27467

000

Zhenjun Zhao @ericzzj.bsky.social · 23/09/2026

Dual Covariance Gaussian Splatting SLAM: Decoupling Rendering and Registration for Robust Real-Time Tracking Edward Beng Wai Tan, Siew-Kei Lam tl;dr: each Gaussian has tracking/mapping covariances shared with mean arxiv.org/abs/2609.25746

Zhenjun Zhao @ericzzj.bsky.social · 23/09/2026

NaCR: Visual Localization via NeRF-aided Camera Ray Regression Yesheng Zhang, Xiang Dai, Xu Zhao, Chongyang Zhang tl;dr: NeRF + CRR arxiv.org/abs/2609.25907

000

Zhenjun Zhao @ericzzj.bsky.social · 23/09/2026

MatcherCompass: A Deployment-Aware Benchmark to Guide Image Matcher Selection in the Wild Hyunwoo Kim, @giseopkim.bsky.social tl;dr: in title arxiv.org/abs/2609.25688

000

RegVGGT: Sustainable Visual Geometry Grounding for Streaming via Regulated Memory Hongbo Mao, Junjun Jiang, Youyu Chen, Jiaxin Zhang, Zhemeng Dong, Xianming Liu tl;dr: retain at most 1% tokens per frame to update the context memory arxiv.org/abs/2609.23286

000

D3GS: Depth, DINO, and RGB Diffusion Co-Guided 3D Gaussian Splatting for Sparse-View Reconstruction Yunqi Gao, Zhanfeng Liao, Hanzhang Tu, Zhaoqi Su, Guoqing Zheng, Songtao Wang, Hongwen Zhang, Zhou Xue, Leyuan Liu, Yebin Liu arxiv.org/abs/2609.22941

110

VGGT-Prime: Compute-Adaptive Mixture-of-Heads for Efficient Visual Geometry Transformers Abteen Arab, Guile Wu, Chengjie Huang, Dongfeng Bai tl;dr: router->three computation modes for each global-attention head->mean pooling/surrogate attention/softmax attention arxiv.org/abs/2609.23733

000

Elevator-VIGS: Separating Elevator Motion from Robot Motion in Visual-Inertial Gaussian Splatting SLAM Rui Zhou, Zihan Zhu, Wei Zhang, Zizhou Luo, Norbert Haala, @marcpollefeys.bsky.social tl;dr: VIGS-SLAM->elevator rides arxiv.org/abs/2609.23491

000

BayesianGS-SLAM: Uncertainty-Aware Neural Rendering SLAM via Probabilistic Formulation Kyeongsu Kang, Seongbo Ha, Sibaek Lee, Hyeonwoo Yu tl;dr: sensor-noise uncertainty + opacity-induced map-representation uncertainty->mapping, tracking, keyframe selection arxiv.org/abs/2609.24140