Filters








10,624 Hits in 3.3 sec

Qualitative Pose Estimation by Discriminative Deformable Part Models [chapter]

Hyungtae Lee, Vlad I. Morariu, Larry S. Davis
2013 Lecture Notes in Computer Science  
Unlike previous approaches, our parts are both selected and trained to improve qualitative pose discrimination and are shared by all the qualitative pose models.  ...  We present a discriminative deformable part model for the recovery of qualitative pose, inferring coarse pose labels (e.g., left, frontright, back), a task which we expect to be more robust to common confounding  ...  [11] 0.2814 1.9431 Our approach 0.3485 1.6810 Conclusions We presented a qualitative pose estimation approach that is based on discriminative deformable part models.  ... 
doi:10.1007/978-3-642-37444-9_57 fatcat:vut55lbhgbfkfjtlpt4sjc2bi4

Multi-scale Attention Guided Pose Transfer [article]

Prasun Roy, Saumik Bhattacharya, Subhankar Ghosh, Umapada Pal
2022 arXiv   pre-print
In this paper, we present an improved network architecture for pose transfer by introducing attention links at every resolution level of the encoder and decoder.  ...  Pose transfer refers to the probabilistic image generation of a person with a previously unseen novel pose from another image of that person having a different pose.  ...  In Fig. 5 , we show a qualitative comparison among the generated images by different model variants. Model A0 generates blurry and visually inconsistent images.  ... 
arXiv:2202.06777v1 fatcat:e3hj2cvhhnadjcx5rrpokhgm44

Human Pose Transfer by Adaptive Hierarchical Deformation [article]

Jinsong Zhang, Xingzi Liu, Kun Li
2020 arXiv   pre-print
the image layer by layer.  ...  In this paper, we propose an adaptive human pose transfer network with two hierarchical deformation levels.  ...  Acknowledgements This work was supported in part by Tianjin Research Program of Application Foundation and Advanced Technology (18JCY-BJC19200).  ... 
arXiv:2012.06940v1 fatcat:v5tiht5l3jhm5hofpake7e25wm

Deformation-aware Unpaired Image Translation for Pose Estimation on Laboratory Animals [article]

Siyuan Li, Semih Günel, Mirela Ostrek, Pavan Ramdya, Pascal Fua, and Helge Rhodin
2020 arXiv   pre-print
Our model lets us train a pose estimator on the target domain by transferring readily available body keypoint locations from the source domain to generated target images.  ...  Human pose estimation attains remarkable accuracy when trained on real or simulated datasets consisting of millions of frames.  ...  Figure 9 . 9 Qualitative pose estimation results. The estimator provides decent results across all three animals.  ... 
arXiv:2001.08601v1 fatcat:ykl667mw4rd5tgth4rg6ms2fbq

A Joint Model for 2D and 3D Pose Estimation from a Single Image

Edgar Simo-Serra, Ariadna Quattoni, Carme Torras, Francesc Moreno-Noguer
2013 2013 IEEE Conference on Computer Vision and Pattern Recognition  
For this purpose, we propose a Bayesian framework that integrates a generative model based on latent variables and discriminative 2D part detectors based on HOGs, and perform inference using evolutionary  ...  Real experimentation demonstrates competitive results, and the ability of our methodology to provide accurate 2D and 3D pose estimations even when the 2D detectors are inaccurate.  ...  We have shown it is not only possible to estimate 3D poses by applying part detectors used in 2D pose estimation, but that it is also beneficial to the 2D pose estimation itself, as the 2D deformations  ... 
doi:10.1109/cvpr.2013.466 dblp:conf/cvpr/Simo-SerraQTM13 fatcat:7g7dy5yuezclxmum7sxgedfmvi

Recognizing actions from still images

Nazli Ikizler, R. Gokberk Cinbis, Selen Pehlivan, Pinar Duygulu
2008 Pattern Recognition (ICPR), Proceedings of the International Conference on  
Our results over a new dataset collected for this problem show that by using a rectangle histogramming approach, we can discriminate actions to a great extent.  ...  Our method involves representing the pose with a spatial and orientational histogramming of rectangular regions on a parse probability map.  ...  The overall edge-based deformable model is used to estimate the initial body part positions.  ... 
doi:10.1109/icpr.2008.4761663 dblp:conf/icpr/IkizlerCPD08 fatcat:iat3gi6lh5d65n2bmwuxakv67a

A learning-based markerless approach for full-body kinematics estimation in-natura from a single image

Ami Drory, Hongdong Li, Richard Hartley
2017 Journal of Biomechanics  
By using a discriminatively learned mixture-of-parts model, we construct a probabilistic tree representation to model the configuration and appearance of human body joints.  ...  In the testing stage, the learned models are employed to recover body pose via searching in a test image over a pyramid structure.  ...  We opt for a discriminative part-based approach that requires an offline learning of a model that recovers pose estimates from observable image metrics.  ... 
doi:10.1016/j.jbiomech.2017.01.028 pmid:28237186 fatcat:lihio54rkbcrzc72jvffikvlny

MonoPerfCap: Human Performance Capture from Monocular Video [article]

Weipeng Xu, Avishek Chatterjee, Michael Zollhöfer, Helge Rhodin, Dushyant Mehta, Hans-Peter Seidel, Christian Theobalt
2018 arXiv   pre-print
We tackle these challenges by a novel approach that employs sparse 2D and 3D human pose detections from a convolutional neural network using a batch-based pose estimation strategy.  ...  Our approach reconstructs articulated human skeleton motion as well as medium-scale non-rigid surface deformations in general scenes.  ...  To alleviate this problem, we propose a surface refinement approach to better align the deformed model with automatically estimated actor silhouettes that are found by a model-guided foreground segmentation  ... 
arXiv:1708.02136v2 fatcat:toewmmbynnbppmxsop43d4xk3e

Inverse Graphics with Probabilistic CAD Models [article]

Tejas D. Kulkarni and Vikash K. Mansinghka and Pushmeet Kohli and Joshua B. Tenenbaum
2014 arXiv   pre-print
We apply this approach to 3D human pose estimation and object shape reconstruction from single images, achieving quantitative and qualitative performance improvements over state-of-the-art baselines.  ...  We show that it is possible to solve challenging, real-world 3D vision problems by approximate inference in generative models for images based on rendering the outputs of probabilistic CAD (PCAD) programs  ...  Tejas Kulkarni was graciously supported by the Henry E. Singleton Fellowship. Partly funded by the DARPA PPAML program, grants from the ONR and ARO, and Google's "Rethinking AI" project.  ... 
arXiv:1407.1339v1 fatcat:m55fb5gwazfvtm3pj3jrcqzeha

PhysXNet: A Customizable Approach for LearningCloth Dynamics on Dressed People [article]

Jordi Sanchez-Riera, Albert Pumarola, Francesc Moreno-Noguer
2021 arXiv   pre-print
conditional GAN with a discriminator that enforces feasible deformations.  ...  PhysXNet, by contrast, is a fully differentiable deep network that at inference is able to estimate the geometry of dense cloth meshes in a matter of milliseconds, and thus, can be readily deployed as  ...  Acknowledgment This work is supported partly by the Spanish government under project MoHuCo PID2020-120049RB-I00, the ERA-Net Chistera project IPALM PCI2019-103386 and María de Maeztu Seal of Excellence  ... 
arXiv:2111.07195v1 fatcat:5ididliywnfivpkoqh3zeovlvm

Learning Hierarchical Models for Class-Specific Reconstruction from Natural Data

Arun CS Kumar, Suchendra M. Bhandarkar, Mukta Prasad
2018 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)  
The pipeline combines data-driven deep learning based semantic part learning with principled modelling and effective optimization of the problem's physics, shape deformation, pose and occlusion.  ...  Then, given a new image, it can estimate camera pose and deformable reconstruction using an effective, incremental optimization.  ...  Algorithm 1 RANSAC-based Pose Estimation Algorithm 4: Estimate the pose and deformation parameters by estimating reprojection loss between the mean shape and the 2D part positions chosen in step 1 (above  ... 
doi:10.1109/cvprw.2018.00153 dblp:conf/cvpr/KumarBP18b fatcat:vdklfd63knbvpkafs6cxjalgzi

Single Image Pop-Up from Discriminatively Learned Parts

Menglong Zhu, Xiaowei Zhou, Kostas Daniilidis
2015 2015 IEEE International Conference on Computer Vision (ICCV)  
Given a training set of view exemplars, we learn and select appearance-based discriminative parts which are mapped onto the 3D model through a facility location optimization.  ...  The training set of 3D models is summarized into a set of basis shapes from which we can generalize by linear combination. Given a test image, we detect hypotheses for each part.  ...  The 2D appearance is modeled as a collection of discriminatively trained parts. Each part is associated with a 3D landmark point on a deformable 3D shape.  ... 
doi:10.1109/iccv.2015.112 dblp:conf/iccv/ZhuZD15 fatcat:sg4f3pcx45bnlp5xcgdln3lbzq

GDFace: Gated Deformation for Multi-View Face Image Synthesis

Xuemiao Xu, Keke Li, Cheng Xu, Shengfeng He
2020 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
However, they fail to consider the internal deformation caused by the change of poses, leading to the unsatisfactory synthesized results for large pose variations.  ...  The first module estimates the deformation of two views in the form of convolution offsets according to the input and target poses.  ...  Acknowledgements The work is supported by NSFC (Grant No. 61772206, U1611461, 61472145, 61702194, 61972162)  ... 
doi:10.1609/aaai.v34i07.6942 fatcat:nipqmmp4g5a5za3h7crunlsice

ACR-Pose: Adversarial Canonical Representation Reconstruction Network for Category Level 6D Object Pose Estimation [article]

Zhaoxin Fan, Zhengbo Song, Jian Xu, Zhicheng Wang, Kejian Wu, Hongyan Liu, Jun He
2021 arXiv   pre-print
Recently, category-level 6D object pose estimation has achieved significant improvements with the development of reconstructing canonical 3D representations.  ...  In this paper, we propose a novel Adversarial Canonical Representation Reconstruction Network named ACR-Pose. ACR-Pose consists of a Reconstructor and a Discriminator.  ...  Then, we present works related to adversarial training. 6D Object Pose Estimation Instance-level 6D object pose estimation only estimates the 6D pose of a particular object can be divided into five parts  ... 
arXiv:2111.10524v1 fatcat:yckswnenwvhwncyqeh2mkeb2qq

Pose and Shape Estimation with Discriminatively Learned Parts [article]

Menglong Zhu, Xiaowei Zhou, Kostas Daniilidis
2015 arXiv   pre-print
Given a training set of view exemplars, we learn and select appearance-based discriminative parts which are mapped onto the 3D model from the training set through a facil- ity location optimization.  ...  We introduce a new approach for estimating the 3D pose and the 3D shape of an object from a single image.  ...  The 2D appearance is modeled as a collection of discriminatively trained parts. Each part is associated with a 3D landmark point on a deformable 3D shape.  ... 
arXiv:1502.00192v1 fatcat:qc752hqvujavffgq447ad7kefq
« Previous Showing results 1 — 15 out of 10,624 results