Filters








5,647 Hits in 6.2 sec

Towards Viewpoint Invariant 3D Human Pose Estimation [article]

Albert Haque, Boya Peng, Zelun Luo, Alexandre Alahi, Serena Yeung, Li Fei-Fei
2016 arXiv   pre-print
We propose a viewpoint invariant model for 3D human pose estimation from a single depth image.  ...  We evaluate our model on a previously published depth dataset and a newly collected human pose dataset containing 100K annotated depth images from extreme viewpoints.  ...  viewpoint invariant 3D human pose estimation.  ... 
arXiv:1603.07076v3 fatcat:hxe6advhrng2vavbx22o6ieriu

Towards Viewpoint Invariant 3D Human Pose Estimation [chapter]

Albert Haque, Boya Peng, Zelun Luo, Alexandre Alahi, Serena Yeung, Li Fei-Fei
2016 Lecture Notes in Computer Science  
We propose a viewpoint invariant model for 3D human pose estimation from a single depth image.  ...  We evaluate our model on a previously published depth dataset and a newly collected human pose dataset containing 100K annotated depth images from extreme viewpoints.  ...  viewpoint invariant 3D human pose estimation.  ... 
doi:10.1007/978-3-319-46448-0_10 fatcat:4jmebq4en5bp3haajxlkxhq5vq

View Invariant 3D Human Pose Estimation [article]

Guoqiang Wei and Cuiling Lan and Wenjun Zeng and Zhibo Chen
2019 arXiv   pre-print
In this paper, we propose a view invariant 3D human pose estimation module to alleviate the effects of viewpoint diversity.  ...  The recent success of deep networks has significantly advanced 3D human pose estimation from 2D images.  ...  View Invariant Hierarchical Correction In real life, 3D human poses present great diversity in viewpoints.  ... 
arXiv:1901.10841v1 fatcat:oj3jaibw5jhszhgotbgwj3r5ti

Picture perception reveals mental geometry of 3D scene inferences

Erin Koch, Famya Baig, Qasim Zaidi
2018 Proceedings of the National Academy of Sciences of the United States of America  
Pose estimation of objects in real scenes is critically important for biological and machine visual systems, but little is known of how humans infer 3D poses from 2D retinal images.  ...  To study how humans estimate 3D poses, we used sticks lying on the ground at 16 equally spaced poses, either alone or joined to another stick at an angle of 45°, 90°, or 135°.  ...  Pose estimation of objects in real scenes is critically important for biological and machine visual systems, but little is known of how humans infer 3D poses from 2D retinal images.  ... 
doi:10.1073/pnas.1804873115 pmid:29987008 fatcat:54w6m4onjvazhhifdfi67f4roe

Learning camera viewpoint using CNN to improve 3D body pose estimation [article]

Mona Fathollahi Ghezelghieh, Rangachar Kasturi, Sudeep Sarkar
2016 arXiv   pre-print
The objective of this work is to estimate 3D human pose from a single RGB image.  ...  In this paper, for the first time, we show that camera viewpoint in combination to 2D joint lo-cations significantly improves 3D pose accuracy without the explicit use of perspective geometry mathematical  ...  Furthermore, we demonstrate that training CNN with synthesized 3D human model makes it invariant to human body shape and clothing texture.  ... 
arXiv:1609.05522v1 fatcat:77uajppkdffwjegpp2zxfgwhg4

Unsupervised View-Invariant Human Posture Representation [article]

Faegheh Sardari, Björn Ommer, Majid Mirmehdi
2021 arXiv   pre-print
To overcome this problem, we present a novel unsupervised approach that learns to extract view-invariant 3D human pose representation from a 2D image without using 3D joint data.  ...  Our model is trained by exploiting the intrinsic view-invariant properties of human pose between simultaneous frames from different viewpoints and their equivariant properties between augmented frames  ...  Our viewinvariant features can be applied directly by downstream tasks to be resilient to human pose variations in unseen viewpoints, unlike unsupervised 3D pose estimation methods such as [23, 24, 25  ... 
arXiv:2109.08730v1 fatcat:7bq3hfft4vd4ngtt6qkbgjiwtq

Multiview-Consistent Semi-Supervised Learning for 3D Human Pose Estimation [article]

Rahul Mitra, Nitesh B. Gundavarapu, Abhishek Sharma, Arjun Jain
2020 arXiv   pre-print
The best performing methods for 3D human pose estimation from monocular images require large amounts of in-the-wild 2D and controlled 3D pose annotated datasets which are costly and require sophisticated  ...  multi-view videos of human motions as additional weak supervision signal to guide 3D human pose regression.  ...  Estimation of 3D human pose and viewpoint from input silhouettes via learning a low dimension manifold is shown in [9] . 2D-pose regression and retrieval by pose similarity embedding is shown in [22,  ... 
arXiv:1908.05293v3 fatcat:35thk5rmcjfwtbfads7ej5mmty

Occlusion-Invariant Rotation-Equivariant Semi-Supervised Depth Based Cross-View Gait Pose Estimation [article]

Xiao Gu, Jianxin Yang, Hanxiao Zhang, Jianing Qiu, Frank Po Wen Lo, Yao Guo, Guang-Zhong Yang, Benny Lo
2021 arXiv   pre-print
Furthermore, depth images of human body exhibit heterogeneous characteristics with viewpoint changes, and the estimated poses under local coordinate systems are expected to go through equivariant rotations  ...  Most existing pose estimation models are sensitive to both issues.  ...  Multi-View Depth Pose Dataset A multi-view dataset with depth data is collected from multiple viewpoints for cross-view human pose estimation benchmarking.  ... 
arXiv:2109.01397v1 fatcat:vnhlupdco5hxpbgczujwxt7m4e

Mental geometry of 3D size and shape perception [article]

Qasim Zaidi, Akihito Maruya
2019 bioRxiv   pre-print
We examine if humans have learned to exploit projective geometry to estimate sizes and shapes of objects in 3D scenes.  ...  Retinal images of 3D objects are mapped by the rules of projective geometry, and preserve the invariants of that geometry.  ...  We examine if humans have learned to exploit projective geometry to estimate sizes and shapes of objects in 3D scenes.  ... 
doi:10.1101/761650 fatcat:z6km3lpip5gjlivotemg5cmgtq

DECA: Deep viewpoint-Equivariant human pose estimation using Capsule Autoencoders [article]

Nicola Garau, Niccolò Bisagno, Piotr Bródka, Nicola Conci
2021 arXiv   pre-print
Human Pose Estimation (HPE) aims at retrieving the 3D position of human joints from images or videos.  ...  We show that current 3D HPE methods suffer a lack of viewpoint equivariance, namely they tend to fail or perform poorly when dealing with viewpoints unseen at training time.  ...  When presented with unseen viewpoints, humans display a remarkable ability to estimate human poses, even in the presence of occlusions and unconventional joints configurations.  ... 
arXiv:2108.08557v1 fatcat:r6eyqvwst5cericbhmpyi4jit4

HUMAN ACTION POSELETS ESTIMATION VIA COLOR G-SURF IN STILL IMAGES

M. Favorskaya, D. Novikov, Y. Savitskaya
2015 The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences  
On the other hand, the spatial information about static human poses is valuable for human action recognition.  ...  The proposed approach was tested on samples from PASCAL Visual Object Classes (VOC) Dataset and Challenge 2010 providing accuracy 61-68% for all possible 3D poses locations and 82-86% for front poses locations  ...  Bourdev and Malik (Bourdev and Malik, 2009) were the first, who formulated the task of human pose estimation and recognition as 3D object representation.  ... 
doi:10.5194/isprsarchives-xl-5-w6-51-2015 fatcat:j5tflaktljcetfwjpyodygeiku

Predicting Camera Viewpoint Improves Cross-dataset Generalization for 3D Human Pose Estimation [article]

Zhe Wang, Daeyun Shin, Charless C. Fowlkes
2020 arXiv   pre-print
Monocular estimation of 3d human pose has attracted increased attention with the availability of large ground-truth motion capture datasets.  ...  Based on this observation, we propose an auxiliary task of predicting the camera viewpoint in addition to pose.  ...  of deep neural networks (CNNs), there are significant improvements on 3D human pose estimation [5, 17, 23, 44] .  ... 
arXiv:2004.03143v1 fatcat:fddvykzqlbd5pk5vxib3rfnxri

Mental geometry of three-dimensional size perception

Akihito Maruya, Qasim Zaidi
2020 Journal of Vision  
supplement the generic viewpoint assumption.  ...  We examine if humans have learned to exploit projective geometry to estimate sizes and aspects of three-dimensional shape that are related to relative lengths and aspect ratios.  ...  Discussion The first empirical contribution of this study is to measure size estimates of 3D objects as they are rotated on a ground plane into different poses, which is equivalent to changing viewpoints  ... 
doi:10.1167/jov.20.8.14 pmid:32766745 pmcid:PMC7438683 fatcat:awn5ekpejrbozixtyh44dup7lm

Inferring Unseen Views of People

Chao-Yeh Chen, Kristen Grauman
2014 2014 IEEE Conference on Computer Vision and Pattern Recognition  
Given images of people organized by their rough viewpoint, we form a 3D appearance tensor indexed by images (pose examples), viewpoints, and image positions.  ...  Furthermore, we demonstrate their value for recognizing actions in unseen views and estimating viewpoint in novel images.  ...  This is to be expected, since humans tend to take photos of other humans as ? Viewpoint Pose instance Figure 2 .  ... 
doi:10.1109/cvpr.2014.258 dblp:conf/cvpr/ChenG14a fatcat:5g4wqre5gfa4bli5umwbuntc6a

Motion-Based View-Invariant Articulated Motion Detection and Pose Estimation Using Sparse Point Features [chapter]

Shrinivas J. Pundlik, Stanley T. Birchfield
2009 Lecture Notes in Computer Science  
This observational probability is fed to a Hidden Markov Model defined over multiple poses and viewpoints to obtain temporally consistent pose estimates.  ...  To estimate the pose and viewpoint we introduce a novel motion descriptor that computes the spatial relationships of motion vectors representing various parts of the person using the trajectories of a  ...  We have presented a motion-based approach for detection, tracking, and pose estimation of articulated human motion that is invariant of scale, viewpoint, illumination, and camera motion.  ... 
doi:10.1007/978-3-642-10331-5_40 fatcat:txwtb4piwfghbk3y7msmtoqrjm
« Previous Showing results 1 — 15 out of 5,647 results