Filters








798 Hits in 7.3 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.  ...  Our approach leverages a convolutional and recurrent network architecture with a top-down error feedback mechanism to self-correct previous pose estimates in an end-to-end manner.  ...  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.  ...  Our approach leverages a convolutional and recurrent network architecture with a top-down error feedback mechanism to self-correct previous pose estimates in an end-to-end manner.  ...  viewpoint invariant 3D human pose estimation.  ... 
doi:10.1007/978-3-319-46448-0_10 fatcat:4jmebq4en5bp3haajxlkxhq5vq

Pose Guided Structured Region Ensemble Network for Cascaded Hand Pose Estimation [article]

Xinghao Chen, Guijin Wang, Hengkai Guo, Cairong Zhang
2018 arXiv   pre-print
Hand pose estimation from a single depth image is an essential topic in computer vision and human computer interaction.  ...  In this paper we propose a Pose guided structured Region Ensemble Network (Pose-REN) to boost the performance of hand pose estimation.  ...  [23] proposed a method to iteratively refine the hand pose using hand-crafted 3D pose index features that are invariant to viewpoint transformation. Oberweger et al.  ... 
arXiv:1708.03416v2 fatcat:rm4e4zp5gnbfhhetb2x67il4zq

V2V-PoseNet: Voxel-to-Voxel Prediction Network for Accurate 3D Hand and Human Pose Estimation from a Single Depth Map

Ju Yong Chang, Gyeongsik Moon, Kyoung Mu Lee
2018 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition  
Our system outperforms previous methods in almost all publicly available 3D hand and human pose estimation datasets and placed first in the HANDS 2017 frame-based 3D hand pose estimation challenge.  ...  To overcome these weaknesses, we firstly cast the 3D hand and human pose estimation problem from a single depth map into a voxel-to-voxel prediction that uses a 3D voxelized grid and estimates the per-voxel  ...  We compared the system with state-of-the-art works, which include RTW [56] , viewpoint-invariant feature-based method [15] , and REN-9x6x6 [13] .  ... 
doi:10.1109/cvpr.2018.00533 dblp:conf/cvpr/MoonCL18 fatcat:m6kskcckxjhehmuu54l256tsaq

V2V-PoseNet: Voxel-to-Voxel Prediction Network for Accurate 3D Hand and Human Pose Estimation from a Single Depth Map [article]

Gyeongsik Moon, Ju Yong Chang, Kyoung Mu Lee
2018 arXiv   pre-print
Our system outperforms previous methods in almost all publicly available 3D hand and human pose estimation datasets and placed first in the HANDS 2017 frame-based 3D hand pose estimation challenge.  ...  To overcome these weaknesses, we firstly cast the 3D hand and human pose estimation problem from a single depth map into a voxel-to-voxel prediction that uses a 3D voxelized grid and estimates the per-voxel  ...  2D CNN 3D CNN Depth-based 3D human pose estimation.  ... 
arXiv:1711.07399v3 fatcat:jtp3hizwunglbcn3erid2pavaa

Towards Good Practices for Deep 3D Hand Pose Estimation [article]

Hengkai Guo, Guijin Wang, Xinghao Chen, Cairong Zhang
2017 arXiv   pre-print
3D hand pose estimation from single depth image is an important and challenging problem for human-computer interaction.  ...  To exploit the good practice and promote the performance for hand pose estimation, we propose a tree-structured Region Ensemble Network (REN) for directly 3D coordinate regression.  ...  [14] introduce a viewpoint invariant model using ConvNets and recurrent networks (RNNs) for human pose estimation.  ... 
arXiv:1707.07248v1 fatcat:x2xb7m2oezakzk3ebtikgu3dsu

Fusing Visual and Inertial Sensors with Semantics for 3D Human Pose Estimation

Andrew Gilbert, Matthew Trumble, Charles Malleson, Adrian Hilton, John Collomosse
2018 International Journal of Computer Vision  
We propose an approach to accurately estimate 3D human pose by fusing multi-viewpoint video (MVV) with inertial measurement unit (IMU) sensor data, without optical markers, a complex hardware setup or  ...  Uniquely we use a multi-channel 3D convolutional neural network to learn a pose embedding from visual occupancy and semantic 2D pose estimates from the MVV in a discretised volumetric probabilistic visual  ...  Human Pose Result Fig. 2 Our two-stream network fuses IMU data with volumetric (PVH) data derived from multiple viewpoint video (MVV) to learn an embedding for 3D joint locations (human pose)  ... 
doi:10.1007/s11263-018-1118-y fatcat:n7k67bu5szaztgwqsft5lhvas4

3D Canonical Pose Estimation and Abnormal Gait Recognition with a Single RGB-D Camera

Yao Guo, Fani Deligianni, Xiao Gu, Guang-Zhong Yang
2019 IEEE Robotics and Automation Letters  
Firstly, view-invariant 3D lower limb pose estimation is achieved by fusing information from depth images along with 2D joints derived in RGB images.  ...  Next, both the 6D camera pose and the 3D lower limb skeleton are real-time tracked in a canonical coordinate system based on Simultaneously Localization and Mapping (SLAM).  ...  Information Fusion for 3D Lower Limb Pose Estimation Real-time, view-invariant, human pose estimation from RGB-D images has attracted much attention in recent years [7] , [11] , [19] , in which they  ... 
doi:10.1109/lra.2019.2928775 fatcat:4etczszm35hjnigssyn3ft5oaq

A Survey on 3D Hand Skeleton and Pose Estimation by Convolutional Neural Network

Van-Hung Le, Hung-Cuong Nguyen
2020 Advances in Science, Technology and Engineering Systems  
Restoring, estimating the fully 3D hand skeleton and pose from the image data of the captured sensors/cameras applied in many applications of computer vision and robotics: human-computer interaction; gesture  ...  The study discussed several areas of 3D hand pose estimation: (i)the number of valuable studies about 3D hand pose estimation, (ii) estimates of 3D hand pose when using 3D CNNs and 2D CNNs, (iii) challenges  ...  The average of 3D joints error is 11mm with the NYU dataset. In the past year, the depth image also is used to estimate 3D hand pose in some studies. Yoo et al.  ... 
doi:10.25046/aj050418 fatcat:tzpjnmpwtjbh7m6ld3nucyvxia

Shape from Contour: Computation and Representation

James H. Elder
2018 Annual Review of Vision Science  
More complete shape representations may involve recurrent processes that integrate local and global cues.  ...  The human visual system reliably extracts shape information from complex natural scenes in spite of noise and fragmentation caused by clutter and occlusions.  ...  In the 3D model-based theory, the brain uses 2D features of the image projection of an object to estimate the 3D geometry and pose of the object, allowing comparison with stored 3D object models (e.g.,  ... 
doi:10.1146/annurev-vision-091517-034110 pmid:30222530 fatcat:mwhcyypwpvehxk5l5i7bc2rkbi

Sketching shiny surfaces

Ulrich Weidenbacher, Pierre Bayerl, Heiko Neumann, Roland Fleming
2006 ACM Transactions on Applied Perception  
In this paper we propose a biologically motivated recurrent model for the extraction of visual features relevant for the perception of 3D shape information from images of mirrored objects.  ...  We qualitatively and quantitatively analyze the results of computational model simulations and show that bidirectional recurrent information processing leads to better results than pure feedforward processing  ...  ill-posed [Hadamard 1902] .  ... 
doi:10.1145/1166087.1166094 fatcat:5ihlngx6evfrjeg4szd77ddmye

Learning Motion-Dependent Appearance for High-Fidelity Rendering of Dynamic Humans from a Single Camera [article]

Jae Shin Yoon, Duygu Ceylan, Tuanfeng Y. Wang, Jingwan Lu, Jimei Yang, Zhixin Shu, Hyun Soo Park
2022 arXiv   pre-print
Our experiments show that our method can generate a temporally coherent video of dynamic humans for unseen body poses and novel views given a single view video.  ...  Appearance of dressed humans undergoes a complex geometric transformation induced not only by the static pose but also by its dynamics, i.e., there exists a number of cloth geometric configurations given  ...  Acknowledgement We would like to thank Julien Philip for providing useful feedback on our paper draft. Jae Shin Yoon is supported by Doctoral Dissertation Fellowship from University of Minnesota.  ... 
arXiv:2203.12780v1 fatcat:i2o7bdrgrrav5ajox6i6nzj3xe

Sim2Real View Invariant Visual Servoing by Recurrent Control [article]

Fereshteh Sadeghi, Alexander Toshev, Eric Jang, Sergey Levine
2017 arXiv   pre-print
In this paper, we study how viewpoint-invariant visual servoing skills can be learned automatically in a robotic manipulation scenario.  ...  Humans are remarkably proficient at controlling their limbs and tools from a wide range of viewpoints and angles, even in the presence of optical distortions.  ...  , in order to train a recurrent controller in simulation for viewpoint invariant visual servoing.  ... 
arXiv:1712.07642v1 fatcat:vh5kc4nxfrfpdavfhla5qlcmgu

Object recognition in 3D lidar data with recurrent neural network

Danil V. Prokhorov
2009 2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops  
The method is illustrated on a two-class problem with real data.  ...  This paper introduces a new method for object recognition which is based on a recurrent neural network trained in a supervised mode.  ...  output-toinput feedback instead of RMLP with its internal feedback is not meaningful for our classification task.  ... 
doi:10.1109/cvprw.2009.5204114 dblp:conf/cvpr/Prokhorov09 fatcat:z2le3vxcvnastpmldvy74tv64a

3D Human Pose Estimation Using Convolutional Neural Networks with 2D Pose Information [article]

Sungheon Park, Jihye Hwang, Nojun Kwak
2016 arXiv   pre-print
While there has been a success in 2D human pose estimation with convolutional neural networks (CNNs), 3D human pose estimation has not been thoroughly studied.  ...  In this paper, we tackle the 3D human pose estimation task with end-to-end learning using CNNs. Relative 3D positions between one joint and the other joints are learned via CNNs.  ...  Different colors are used to distinguish the left and right sides of human bodies. It can be found that 2D pose estimation results help reducing the error of 3D pose estimation.  ... 
arXiv:1608.03075v2 fatcat:525yyz6q3veqrobabgo6hqm6gu
« Previous Showing results 1 — 15 out of 798 results