Filters








6,157 Hits in 5.7 sec

Feature Mapping for Learning Fast and Accurate 3D Pose Inference from Synthetic Images [article]

Mahdi Rad, Markus Oberweger, Vincent Lepetit
2018 arXiv   pre-print
the features for the image, map them to the feature space of synthetic images, and finally use the resulting features as input to another network which predicts the 3D pose.  ...  We demonstrate our approach on the LINEMOD dataset for 3D object pose estimation from color images, and the NYU dataset for 3D hand pose estimation from depth maps.  ...  Acknowledgment This work was funded by the Christian Doppler Laboratory for Semantic 3D Computer Vision. We thank Alexander Grabner for helpful discussions.  ... 
arXiv:1712.03904v2 fatcat:2de2hp4d5vhmlbt62qkyxphg7u

Feature Mapping for Learning Fast and Accurate 3D Pose Inference from Synthetic Images

Mahdi Rad, Markus Oberweger, Vincent Lepetit
2018 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition  
the features for the image, map them to the feature space of synthetic images, and finally use the resulting features as input to another network which predicts the 3D pose.  ...  We demonstrate our approach on the LINEMOD dataset for 3D object pose estimation from color images, and the NYU dataset for 3D hand pose estimation from depth maps.  ...  Acknowledgment This work was funded by the Christian Doppler Laboratory for Semantic 3D Computer Vision. We thank Alexander Grabner for helpful discussions.  ... 
doi:10.1109/cvpr.2018.00490 dblp:conf/cvpr/RadOL18 fatcat:l6q2tbhn5vgetlaavfdrr5ovge

Inferring 3D structure with a statistical image-based shape model

Grauman, Shakhnarovich, Darrell
2003 Proceedings Ninth IEEE International Conference on Computer Vision  
Using a dataset of thousands of pedestrian images generated from a synthetic model, we can perform accurate inference of the 3D locations of 19 joints on the body based on observed silhouette contours  ...  Model matching and parameter inference are done entirely in the image domain and require no explicit 3D construction.  ...  We learn an implicit, image-based representation of a known 3D shape, match it to input images using a statistical model, and infer 3D parameters from the matched model.  ... 
doi:10.1109/iccv.2003.1238408 dblp:conf/iccv/GraumanSD03 fatcat:a3a2g3ubxzfcfp3hl4p7lo45jq

MobilePose: Real-Time Pose Estimation for Unseen Objects with Weak Shape Supervision [article]

Tingbo Hou, Adel Ahmadyan, Liangkai Zhang, Jianing Wei, Matthias Grundmann
2020 arXiv   pre-print
In this paper, we address the problem of detecting unseen objects from RGB images and estimating their poses in 3D. We propose two mobile friendly networks: MobilePose-Base and MobilePose-Shape.  ...  Consequently, we add shape prediction as an intermediate layer in the MobilePose-Shape, and let the network learn pose from shape.  ...  We show that our model can infer accurate poses from weakly learned coarse shape features.  ... 
arXiv:2003.03522v1 fatcat:h4a2fgbnxrh7phljsqimaeyluy

Real-time human pose recognition in parts from single depth images

Jamie Shotton, Andrew Fitzgibbon, Mat Cook, Toby Sharp, Mark Finocchio, Richard Moore, Alex Kipman, Andrew Blake
2011 CVPR 2011  
We propose a new method to quickly and accurately predict 3D positions of body joints from a single depth image, using no temporal information.  ...  Our evaluation shows high accuracy on both synthetic and real test sets, and investigates the effect of several training parameters.  ...  We also thank John Winn, Duncan Robertson, Antonio Criminisi, Shahram Izadi, Ollie Williams, and Mihai Budiu for help and valuable discussions, and Varun Ganapathi and Christian Plagemann for providing  ... 
doi:10.1109/cvpr.2011.5995316 dblp:conf/cvpr/ShottonFCSFMKB11 fatcat:nwq7ufwe2fcihjjr2w4f36rmg4

Real-Time Human Pose Recognition in Parts from Single Depth Images [chapter]

Jamie Shotton, Andrew Fitzgibbon, Mat Cook, Toby Sharp, Mark Finocchio, Richard Moore, Alex Kipman, Andrew Blake
2013 Studies in Computational Intelligence  
We propose a new method to quickly and accurately predict 3D positions of body joints from a single depth image, using no temporal information.  ...  Our evaluation shows high accuracy on both synthetic and real test sets, and investigates the effect of several training parameters.  ...  We also thank John Winn, Duncan Robertson, Antonio Criminisi, Shahram Izadi, Ollie Williams, and Mihai Budiu for help and valuable discussions, and Varun Ganapathi and Christian Plagemann for providing  ... 
doi:10.1007/978-3-642-28661-2_5 fatcat:ltt4bcflfnbdrd6yjqa7vdskmi

A Comprehensive Study on Deep Learning-Based 3D Hand Pose Estimation Methods

Theocharis Chatzis, Andreas Stergioulas, Dimitrios Konstantinidis, Kosmas Dimitropoulos, Petros Daras
2020 Applied Sciences  
Nonetheless, finger occlusions and rapid motions still pose significant challenges to the accuracy of such methods.  ...  The field of 3D hand pose estimation has been gaining a lot of attention recently, due to its significance in several applications that require human-computer interaction (HCI).  ...  Accordingly, they trained a network to minimize the distance between the mapped and the synthetic features and a decoder trained on the synthetic images to infer the 3D pose.  ... 
doi:10.3390/app10196850 fatcat:hgyqkoyetbbilncksguarqz3bq

Real-time human pose recognition in parts from single depth images

Jamie Shotton, Toby Sharp, Alex Kipman, Andrew Fitzgibbon, Mark Finocchio, Andrew Blake, Mat Cook, Richard Moore
2013 Communications of the ACM  
We propose a new method to quickly and accurately predict 3D positions of body joints from a single depth image, using no temporal information.  ...  Our evaluation shows high accuracy on both synthetic and real test sets, and investigates the effect of several training parameters.  ...  We also thank John Winn, Duncan Robertson, Antonio Criminisi, Shahram Izadi, Ollie Williams, and Mihai Budiu for help and valuable discussions, and Varun Ganapathi and Christian Plagemann for providing  ... 
doi:10.1145/2398356.2398381 fatcat:ov3jdqq4ufhpjlkvyylwlsruoi

Sparse probabilistic regression for activity-independent human pose inference

Raquel Urtasun, Trevor Darrell
2008 2008 IEEE Conference on Computer Vision and Pattern Recognition  
We report results on synthetic (Poser) and real (Humaneva) pose databases, obtaining fast and accurate pose estimates using training set sizes up to 10 5 .  ...  We propose an online probabilistic regression scheme for efficient inference of complex, highdimensional, and multimodal mappings.  ...  Acknowledgments We thank Robert Wang for generating the Poser Hand Database and the authors of [7] for sharing their executable to compute the Hierarchical features.  ... 
doi:10.1109/cvpr.2008.4587360 dblp:conf/cvpr/UrtasunD08 fatcat:4wgpdvyernghdangxcvhvvolym

Real-Time Eye Tracking for Bare and Sunglasses-wearing Faces for Augmented Reality 3D Head-Up Displays

Dongwoo Kang, Lin Ma
2021 IEEE Access  
a fast, accurate, and robust manner.  ...  FAST AND ACCURATE EYE TRACKING WITH IRIS REGRESSION ON BARE FACES For bare faces, we use a coarse-to-fine strategy to infer the pupil center.  ... 
doi:10.1109/access.2021.3110644 fatcat:hresfwg3sndvpdahsscqtj2r6a

Im2Fit: Fast 3D Model Fitting and Anthropometrics using Single Consumer Depth Camera and Synthetic Data [article]

Qiaosong Wang, Vignesh Jagadeesh, Bryan Ressler, Robinson Piramuthu
2014 arXiv   pre-print
Compared to existing methods, our approach is able to predict accurate full body parameters from a partial view using measurement parameters learned from the synthetic dataset.  ...  In this paper, we propose a method for capturing accurate human body shape and anthropometrics from a single consumer grade depth sensor.  ...  It learns a pose deformation model from 71 registered meshes of a single subject in various poses. However, the human shape model is learned from different subjects with a standard pose.  ... 
arXiv:1410.0745v2 fatcat:mcgzkwby2jekjbxt67t7tdtpve

A Survey on Deep Learning Based Methods and Datasets for Monocular 3D Object Detection

Seong-heum Kim, Youngbae Hwang
2021 Electronics  
Owing to recent advancements in deep learning methods and relevant databases, it is becoming increasingly easier to recognize 3D objects using only RGB images from single viewpoints.  ...  Based on this simple sensor modality for practical applications, deep learning-based monocular 3D object detection methods that overcome significant research challenges are categorized and summarized.  ...  Acknowledgments: The first author sincerely appreciates Min-ho Lee and Hoo-kyeong Lee at KETI for valuable discussion. Conflicts of Interest: The authors declare no conflict of interest.  ... 
doi:10.3390/electronics10040517 fatcat:rziqhrkefvelpg3vgb6qxfprte

Reconstruct Locally, Localize Globally: A Model Free Method for Object Pose Estimation

Ming Cai, Ian Reid
2020 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)  
At inference time, our method maps from the RoI features of the input image to a dense collection of object-centric 3D coordinates, one per pixel.  ...  We show that this method eliminates the requirement for a 3D CAD model (needed by classical geometry-based methods and state-of-the-art learning based methods alike) but still achieves performance on a  ...  Acknowledgement We gratefully acknowledge the support of the Australian Research Council through the Centre of Excellence for Robotic Vision CE140100016 and Laureate Fellowship FL130100102 to IR.  ... 
doi:10.1109/cvpr42600.2020.00322 dblp:conf/cvpr/CaiR20 fatcat:xpc2ths3bff5jbvam4ksbl6sk4

Object Proposals Estimation in Depth Image Using Compact 3D Shape Manifolds [chapter]

Shuai Zheng, Victor Adrian Prisacariu, Melinos Averkiou, Ming-Ming Cheng, Niloy J. Mitra, Jamie Shotton, Philip H. S. Torr, Carsten Rother
2015 Lecture Notes in Computer Science  
Additionally, to support fast and accurate inference, we improve the standard 3D object category proposal generation pipeline by applying a shallow convolutional neural network-based filtering stage.  ...  In this work we investigate the task of finding particular object shapes from a single depth image.  ...  Next we generate 3D shapes back from these spaces and finally render them into multiple 2.5D depth-only projections. A second requirement for a proposal generator is fast and accurate inference.  ... 
doi:10.1007/978-3-319-24947-6_16 fatcat:64e46fknhzfoldu2zgtmsr2o24

Fast Monocular Hand Pose Estimation on Embedded Systems [article]

Shan An, Xiajie Zhang, Dong Wei, Haogang Zhu, Jianyu Yang, Konstantinos A. Tsintotas
2021 arXiv   pre-print
This paper proposes a fast and accurate framework for hand pose estimation, dubbed as "FastHand".  ...  However, previous approaches suffer from unsatisfying hand landmark predictions in real-world scenes and high computation burden.  ...  learning of a 3D hand pose.  ... 
arXiv:2102.07067v3 fatcat:xvtxot3csjextplw6yvqhkflbq
« Previous Showing results 1 — 15 out of 6,157 results