Egogaussian: Dynamic scene understanding from egocentric video with 3d gaussian splatting

D Zhang, G Li, J Li, M Bressieux, O Hilliges… - ar**
Y Ji, H Zhu, J Tang, W Liu, Z Zhang, Y **e… - arxiv preprint arxiv …, 2024 - arxiv.org
The semantically interactive radiance field has always been an appealing task for its
potential to facilitate user-friendly and automated real-world 3D scene understanding …

3D-Aware Instance Segmentation and Tracking in Egocentric Videos

Y Bhalgat, V Tschernezki, I Laina… - Proceedings of the …, 2024 - openaccess.thecvf.com
Egocentric videos present unique challenges for 3D scene understanding due to rapid
camera motion, frequent object occlusions, and limited object visibility. This paper …

SLGaussian: Fast Language Gaussian Splatting in Sparse Views

K Chen, BQ Dai, M Qin, D Zhang, P Li, Y Zou… - arxiv preprint arxiv …, 2024 - arxiv.org
3D semantic field learning is crucial for applications like autonomous navigation, AR/VR,
and robotics, where accurate comprehension of 3D scenes from limited viewpoints is …

Estimating Body and Hand Motion in an Ego-sensed World

B Yi, V Ye, M Zheng, L Müller, G Pavlakos, Y Ma… - arxiv preprint arxiv …, 2024 - arxiv.org
We present EgoAllo, a system for human motion estimation from a head-mounted device.
Using only egocentric SLAM poses and images, EgoAllo guides sampling from a conditional …

SAMa: Material-aware 3D selection and segmentation

M Fischer, I Georgiev, T Groueix, VG Kim… - arxiv preprint arxiv …, 2024 - arxiv.org
Decomposing 3D assets into material parts is a common task for artists and creators, yet
remains a highly manual process. In this work, we introduce Select Any Material (SAMa), a …

T-3DGS: Removing Transient Objects for 3D Scene Reconstruction

V Pryadilshchikov, A Markin, A Komarichev… - arxiv preprint arxiv …, 2024 - arxiv.org
We propose a novel framework to remove transient objects from input videos for 3D scene
reconstruction using Gaussian Splatting. Our framework consists of the following steps. In …

3D Gaussian Splatting for Human-Robot Interaction

S Bowser, SM Lukin - The 1st InterAI Workshop: Interactive AI for …, 2024 - openreview.net
In order to assist a humans' ability to make decisions in uncertain and high-stakes scenarios,
eg, disaster relief, we aim to provide an interactive and “smart” visual model of an …

EgoMe: Follow Me via Egocentric View in Real World

H Qiu, Z Shi, L Wang, H **ong, X Li, H Li - arxiv preprint arxiv:2501.19061, 2025 - arxiv.org
When interacting with the real world, human often take the egocentric (first-person) view as a
benchmark, naturally transferring behaviors observed from a exocentric (third-person) view …

3D Gaussian Splatting in Robotics: A Survey

S Zhu, G Wang, D Kong, H Wang - arxiv preprint arxiv:2410.12262, 2024 - arxiv.org
Dense 3D representations of the environment have been a long-term goal in the robotics
field. While previous Neural Radiance Fields (NeRF) representation have been prevalent for …