Filters








429 Hits in 4.3 sec

VIBE: Video Inference for Human Body Pose and Shape Estimation [article]

Muhammed Kocabas, Nikos Athanasiou, Michael J. Black
2020 arXiv   pre-print
To address this problem, we propose Video Inference for Body Pose and Shape Estimation (VIBE), which makes use of an existing large-scale motion capture dataset (AMASS) together with unpaired, in-the-wild  ...  Despite progress on single-image 3D pose and shape estimation, existing video-based state-of-the-art methods fail to produce accurate and natural motion sequences due to a lack of ground-truth 3D motion  ...  This research was partially supported by the Max Planck ETH Center for Learning Systems and the Max Planck Graduate Center for Computer and Information Science.  ... 
arXiv:1912.05656v3 fatcat:vrhvro6q65ftzpuzehb4ab24bu

VIBE: Video Inference for Human Body Pose and Shape Estimation

Muhammed Kocabas, Nikos Athanasiou, Michael J. Black
2020 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)  
Figure 1 : Given challenging in-the-wild videos, a recent state-of-the-art video-pose-estimation approach [30] (top), fails to produce accurate 3D body poses.  ...  Our model (VIBE) (bottom) is able to produce realistic and accurate pose and shape, outperforming previous work on standard benchmarks.  ...  This research was partially supported by the Max Planck ETH Center for Learning Systems and the Max Planck Graduate Center for Computer and Information Science.  ... 
doi:10.1109/cvpr42600.2020.00530 dblp:conf/cvpr/KocabasAB20 fatcat:ooyp57g76vbppfl67zryxnj47a

Deep Two-Stream Video Inference for Human Body Pose and Shape Estimation [article]

Ziwen Li, Bo Xu, Han Huang, Cheng Lu, Yandong Guo
2021 arXiv   pre-print
In this paper, we propose a new framework Deep Two-Stream Video Inference for Human Body Pose and Shape Estimation (DTS-VIBE), to generate 3D human pose and mesh from RGB videos.  ...  Several video-based 3D pose and shape estimation algorithms have been proposed to resolve the temporal inconsistency of single-image-based methods.  ...  Architecture In this section, we describe the deep two-stream video inference network for human body pose and shape estimation (DTS-VIBE).  ... 
arXiv:2110.11680v1 fatcat:hizuzxdfpzep5dnvy2e24fan6e

ViBE: Dressing for Diverse Body Shapes

Wei-Lin Hsiao, Kristen Grauman
2020 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)  
These body-agnostic vision methods and datasets are a barrier to inclusion, ill-equipped to provide good suggestions for diverse body shapes.  ...  We introduce ViBE, a VIsual Body-aware Embedding that captures clothing's affinity with different body shapes.  ...  Acknowledgements: We thank our human subjects: Angel, Chelsea, Cindy, Layla, MongChi, Ping, Yenyen, and our anonymous friends and volunteers from Facebook.  ... 
doi:10.1109/cvpr42600.2020.01107 dblp:conf/cvpr/HsiaoG20 fatcat:bbnpyp2oljet5hpknhbftnqzaa

ViBE: Dressing for Diverse Body Shapes [article]

Wei-Lin Hsiao, Kristen Grauman
2020 arXiv   pre-print
These body-agnostic vision methods and datasets are a barrier to inclusion, ill-equipped to provide good suggestions for diverse body shapes.  ...  We introduce ViBE, a VIsual Body-aware Embedding that captures clothing's affinity with different body shapes.  ...  Acknowledgements: We thank our human subjects: Angel, Chelsea, Cindy, Layla, MongChi, Ping, Yenyen, and our anonymous friends and volunteers from Facebook.  ... 
arXiv:1912.06697v2 fatcat:tqmvjc36rfa6joto6jkl5obkmy

EventHPE: Event-based 3D Human Pose and Shape Estimation [article]

Shihao Zou, Chuan Guo, Xinxin Zuo, Sen Wang, Pengyu Wang, Xiaoqin Hu, Shoushun Chen, Minglun Gong, Li Cheng
2021 arXiv   pre-print
Event camera is an emerging imaging sensor for capturing dynamics of moving objects as events, which motivates our work in estimating 3D human pose and shape from the event signals.  ...  Both events and optical flow are closely related to human body dynamics, which are fed as input to the ShapeNet in the second stage, to estimate 3D human shapes.  ...  Acknowledgement Thank all the volunteers who contribute to the dataset, and thank Shuang Wu and Wei Ji for their constructive advice.  ... 
arXiv:2108.06819v1 fatcat:yhlkmjbg2fdafkinl5uo5bi3pa

Live Stream Temporally Embedded 3D Human Body Pose and Shape Estimation [article]

Zhouping Wang, Sarah Ostadabbas
2022 arXiv   pre-print
3D Human body pose and shape estimation within a temporal sequence can be quite critical for understanding human behavior.  ...  To address this problem, we present a temporally embedded 3D human body pose and shape estimation (TePose) method to improve the accuracy and temporal consistency of pose estimation in live stream videos  ...  At the start of each live stream or video, we utilize the VIBE [24] , which is a video-based inference model for body pose and shape estimation, to provide the SMPL parameters for the first T frames.  ... 
arXiv:2207.12537v1 fatcat:w2bjgtpecrchzlp6dbrfgn4moq

Camera Motion Agnostic 3D Human Pose Estimation [article]

Seong Hyun Kim, Sunwon Jeong, Sungbum Park, Ju Yong Chang
2021 arXiv   pre-print
This makes it difficult to estimate a person's pure pose and motion in world coordinate system for a video captured using a moving camera.  ...  Although the performance of 3D human pose and shape estimation methods has improved significantly in recent years, existing approaches typically generate 3D poses defined in camera or human-centered coordinate  ...  Related Works Methods for simultaneously reconstructing 3D human poses and shapes are reviewed in this section. 3D human pose and shape estimation from a single image.  ... 
arXiv:2112.00343v1 fatcat:k56y2we2gjgshk7e22weznpvt4

3D Human Motion Estimation via Motion Compression and Refinement [article]

Zhengyi Luo, S. Alireza Golestaneh, Kris M. Kitani
2020 arXiv   pre-print
We develop a technique for generating smooth and accurate 3D human pose and motion estimates from RGB video sequences.  ...  Experiments show that our method produces both smooth and accurate 3D human pose and motion estimates.  ...  Acknowledgements: This project was sponsored in part by IARPA (D17PC00340), and JST AIP Acceleration Research Grant (JPMJCR20U1).  ... 
arXiv:2008.03789v2 fatcat:y3rx7wgwbfeo3ijddukftza6ri

4D Human Body Capture from Egocentric Video via 3D Scene Grounding [article]

Miao Liu, Dexin Yang, Yan Zhang, Zhaopeng Cui, James M. Rehg, Siyu Tang
2021 arXiv   pre-print
Moreover, we compare our method with the previous state-of-the-art method on human motion capture from monocular video, and show that our method estimates more accurate human-body poses and shapes under  ...  The unique viewpoint and rapid embodied camera motion of egocentric videos raise additional technical barriers for human body capture.  ...  More closely-related to this work are prior efforts that leverage video of a moving peprson to infer a time series of 3D human body poses and shapes. Alldieck et al.  ... 
arXiv:2011.13341v2 fatcat:c4a36zej4zhqdajwr4bp2mxo3e

Learning Local Recurrent Models for Human Mesh Recovery [article]

Runze Li and Srikrishna Karanam and Ren Li and Terrence Chen and Bir Bhanu and Ziyan Wu
2021 arXiv   pre-print
We consider the problem of estimating frame-level full human body meshes given a video of a person with natural motion dynamics.  ...  To address these issues, we present a new method for video mesh recovery that divides the human mesh into several local parts following the standard skeletal model.  ...  We presented results of an extensive set of experiments on various challenging benchmark datasets to demonstrate the efficacy of the proposed local recurrent modeling approach to video human mesh recovery  ... 
arXiv:2107.12847v1 fatcat:teypmwsuhbgm3jhkdgl7kltjam

Multi-View Large Population Gait Database With Human Meshes and Its Performance Evaluation

Xiang Li, Yasushi Makihara, Chi Xu, Yasushi Yagi
2022 IEEE Transactions on Biometrics Behavior and Identity Science  
In this paper, we consider a more informative 3D human mesh model with parametric pose and shape features, and propose a multi-view training framework for accurate mesh estimation.  ...  Existing model-based gait databases provide the 2D poses (i.e., joint locations) extracted by general pose estimators as the human model.  ...  ACKNOWLEDGMENT The authors thank Stuart Jenkinson, PhD, from Edanz (https://jp.edanz.com/ac) for editing a draft of this manuscript.  ... 
doi:10.1109/tbiom.2022.3174559 fatcat:vo6vkigozze7hfb36bcq7iji7q

Beyond Static Features for Temporally Consistent 3D Human Pose and Shape from a Video [article]

Hongsuk Choi, Gyeongsik Moon, Ju Yong Chang, Kyoung Mu Lee
2021 arXiv   pre-print
Despite the recent success of single image-based 3D human pose and shape estimation methods, recovering temporally consistent and smooth 3D human motion from a video is still challenging.  ...  Our TCMR significantly outperforms previous video-based methods in temporal consistency with better per-frame 3D pose and shape accuracy. We also release the codes.  ...  This work was supported by IITP grant funded by the Ministry of Science and ICT of Korea  ... 
arXiv:2011.08627v4 fatcat:pkjq5w3wrvh6zmzh3hthxwhdye

Trajectory Optimization for Physics-Based Reconstruction of 3d Human Pose from Monocular Video [article]

Erik Gärtner, Mykhaylo Andriluka, Hongyi Xu, Cristian Sminchisescu
2022 arXiv   pre-print
We focus on the task of estimating a physically plausible articulated human motion from monocular video.  ...  body and scene geometry.  ...  We would like to thank Erwin Coumans for his help with the project, as well as the supportive anonymous reviewers for their insightful comments.  ... 
arXiv:2205.12292v1 fatcat:2vckadaqgrcmpbq35ra2ybtjfy

Imposing Temporal Consistency on Deep Monocular Body Shape and Pose Estimation [article]

Alexandra Zimmer, Anna Hilsmann, Wieland Morgenstern, Peter Eisert
2022 arXiv   pre-print
In extensive experiments, we show that our approach results in accurately estimated body shape and motion, also for challenging movements and poses.  ...  In detail, we derive parameters of a sequence of body models, representing shape and motion of a person, including jaw poses, facial expressions, and finger poses.  ...  The employed data set MoVi is available under [6] , the data set own-data and the code produced for this paper are not publicly available.  ... 
arXiv:2202.03074v2 fatcat:wcsusebynjbdxgkykl5rhoq4sy
« Previous Showing results 1 — 15 out of 429 results