activity
20242026
collaborators

11 papers

cs.CV2026

SplatGuide: Geometric Priors from 3D Gaussians for Pose-Free Novel View Synthesis

Yejun Zhang, Zihan Wang, Xu Ji +8

Generating photorealistic novel views from unposed images requires both 3D geometric understanding and the ability to synthesize unseen content. A natural strategy combines feed-fo…

cs.CV2026

GeoMix: Descriptor-Free Visual Localization via Global Context and Multi-Detector Training

Yejun Zhang, Xinjue Wang, Zihan Wang +2

Descriptor-free visual localization eliminates high-dimensional descriptor storage, preserves scene privacy, and simplifies map maintenance, yet its accuracy still lags far behind…

cs.GR2026

VolHuMe: a High-Resolution Large Scale Dataset of Volumetric Human Meshes

Giulia Martinelli, Niccolò Bisagno, Nicola Garau +2

We introduce VolHuMe, a dataset of high-quality 4D human scans captured with a state-of-the-art volumetric studio using 64 RGB and 32 depth cameras. VolHuMe contains individual cap…

cs.CV2026

3D Gaussian Splatting with Fisheye Images: Field of View Analysis and Depth-Based Initialization

Ulas Gunes, Matias Turkulainen, Mikhail Silaev +2

We present the first evaluation of 3D Gaussian Splatting methods on real fisheye imagery with fields of view above 180\textdegree{}. Our study evaluates Fisheye-GS \cite{liao2024fi…

cs.CV2025

The Weighting Game: Evaluating Quality of Explainability Methods

Lassi Raatikainen, Esa Rahtu

The objective of this paper is to assess the quality of explanation heatmaps for image classification tasks. To assess the quality of explainability methods, we approach the task t…

cs.CV2025

FIORD: A Fisheye Indoor-Outdoor Dataset with LIDAR Ground Truth for 3D Scene Reconstruction and Benchmarking

Ulas Gunes, Matias Turkulainen, Xuqian Ren +3

The development of large-scale 3D scene reconstruction and novel view synthesis methods mostly rely on datasets comprising perspective images with narrow fields of view (FoV). Whil…