Zobrazeno 1 - 10
of 2 348
pro vyhledávání: '"Schwing, A."'
Autor:
Tang, Zhenggang, Fan, Yuchen, Wang, Dilin, Xu, Hongyu, Ranjan, Rakesh, Schwing, Alexander, Yan, Zhicheng
Recent sparse multi-view scene reconstruction advances like DUSt3R and MASt3R no longer require camera calibration and camera pose estimation. However, they only process a pair of views at a time to infer pixel-aligned pointmaps. When dealing with mo
Externí odkaz:
http://arxiv.org/abs/2412.06974
We present RELOCATE, a simple training-free baseline designed to perform the challenging task of visual query localization in long videos. To eliminate the need for task-specific training and efficiently handle long videos, RELOCATE leverages a regio
Externí odkaz:
http://arxiv.org/abs/2412.01826
Decision Transformers have recently emerged as a new and compelling paradigm for offline Reinforcement Learning (RL), completing a trajectory in an autoregressive way. While improvements have been made to overcome initial shortcomings, online finetun
Externí odkaz:
http://arxiv.org/abs/2410.24108
Recent work studying the generalization of diffusion models with UNet-based denoisers reveals inductive biases that can be expressed via geometry-adaptive harmonic bases. However, in practice, more recent denoising networks are often based on transfo
Externí odkaz:
http://arxiv.org/abs/2410.21273
Autor:
Tang, Zhenggang, Zhuang, Peiye, Wang, Chaoyang, Siarohin, Aliaksandr, Kant, Yash, Schwing, Alexander, Tulyakov, Sergey, Lee, Hsin-Ying
The task of image-to-multi-view generation refers to generating novel views of an instance from a single image. Recent methods achieve this by extending text-to-image latent diffusion models to multi-view version, which contains an VAE image encoder
Externí odkaz:
http://arxiv.org/abs/2408.14016
Publikováno v:
2014 IEEE Sensors Applications Symposium (SAS), Queenstown, New Zealand, 2014, pp. 242-247
The frequency dependence of dielectric material properties of water saturated and unsaturated porous materials such as soil is not only disturbing in applications with high frequency electromagnetic (HF-EM) techniques but also contains valuable infor
Externí odkaz:
http://arxiv.org/abs/2406.15756
Autor:
Tang, Zhenggang, Ren, Zhongzheng, Zhao, Xiaoming, Wen, Bowen, Tremblay, Jonathan, Birchfield, Stan, Schwing, Alexander
We present a method for automatically modifying a NeRF representation based on a single observation of a non-rigid transformed version of the original scene. Our method defines the transformation as a 3D flow, specifically as a weighted linear blendi
Externí odkaz:
http://arxiv.org/abs/2406.10543
Precise knowledge of the frequency dependent electromagnetic properties of porous media is urgently necessary for successful utilization of high frequency electromagnetic measurement techniques for near and subsurface sensing. Thus, there is a need o
Externí odkaz:
http://arxiv.org/abs/2406.06789
We introduce GoMAvatar, a novel approach for real-time, memory-efficient, high-quality animatable human modeling. GoMAvatar takes as input a single monocular video to create a digital avatar capable of re-articulation in new poses and real-time rende
Externí odkaz:
http://arxiv.org/abs/2404.07991
We propose the new task 'open-world video instance segmentation and captioning'. It requires to detect, segment, track and describe with rich captions never before seen objects. This challenging task can be addressed by developing "abstractors" which
Externí odkaz:
http://arxiv.org/abs/2404.03657