Pose Machines :. Wei, S., Ramakrishna, V., Kanada, T., & Sheikh, Y. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2016. ![Pose Machines : [pdf]](https://bibbase.org/img/filetypes/pdf.svg "pdf")
Paper abstract bibtex Pose Machines provide a sequential prediction frame-work for learning rich implicit spatial models. In this work we show a systematic design for how convolutional net-works can be incorporated into the pose machine frame-work for learning image features and image-dependent spa-tial models for the task of pose estimation. The contribution of this paper is to implicitly model long-range dependen-cies between variables in structured prediction tasks such as articulated pose estimation. We achieve this by designing a sequential architecture composed of convolutional net-works that directly operate on belief maps from previous stages, producing increasingly refined estimates for part lo-cations, without the need for explicit graphical model-style inference. Our approach addresses the characteristic diffi-culty of vanishing gradients during training by providing a natural learning objective function that enforces intermedi-ate supervision, thereby replenishing back-propagated gra-dients and conditioning the learning procedure. We demon-strate state-of-the-art performance and outperform compet-ing methods on standard benchmarks including the MPII, LSP, and FLIC datasets.
@article{
title = {Pose Machines :},
type = {article},
year = {2016},
id = {5ccfc8d8-0bf4-334c-a755-dc378ba6f7c0},
created = {2022-09-19T10:49:11.819Z},
file_attached = {true},
profile_id = {276016a7-2c9d-3507-8888-093db7c54774},
group_id = {5ec9cc91-a5d6-3de5-82f3-3ef3d98a89c1},
last_modified = {2022-09-19T10:50:34.378Z},
read = {false},
starred = {false},
authored = {false},
confirmed = {true},
hidden = {false},
folder_uuids = {02fb5526-03ff-44ad-8d5c-42bd496c3100},
private_publication = {false},
abstract = {Pose Machines provide a sequential prediction frame-work for learning rich implicit spatial models. In this work we show a systematic design for how convolutional net-works can be incorporated into the pose machine frame-work for learning image features and image-dependent spa-tial models for the task of pose estimation. The contribution of this paper is to implicitly model long-range dependen-cies between variables in structured prediction tasks such as articulated pose estimation. We achieve this by designing a sequential architecture composed of convolutional net-works that directly operate on belief maps from previous stages, producing increasingly refined estimates for part lo-cations, without the need for explicit graphical model-style inference. Our approach addresses the characteristic diffi-culty of vanishing gradients during training by providing a natural learning objective function that enforces intermedi-ate supervision, thereby replenishing back-propagated gra-dients and conditioning the learning procedure. We demon-strate state-of-the-art performance and outperform compet-ing methods on standard benchmarks including the MPII, LSP, and FLIC datasets.},
bibtype = {article},
author = {Wei, Shih-En and Ramakrishna, Varun and Kanada, Takeo and Sheikh, Yaser},
journal = {Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition}
}
Downloads: 0
{"_id":"Ms2LjGREmzuLkWPQb","bibbaseid":"wei-ramakrishna-kanada-sheikh-posemachines-2016","author_short":["Wei, S.","Ramakrishna, V.","Kanada, T.","Sheikh, Y."],"bibdata":{"title":"Pose Machines :","type":"article","year":"2016","id":"5ccfc8d8-0bf4-334c-a755-dc378ba6f7c0","created":"2022-09-19T10:49:11.819Z","file_attached":"true","profile_id":"276016a7-2c9d-3507-8888-093db7c54774","group_id":"5ec9cc91-a5d6-3de5-82f3-3ef3d98a89c1","last_modified":"2022-09-19T10:50:34.378Z","read":false,"starred":false,"authored":false,"confirmed":"true","hidden":false,"folder_uuids":"02fb5526-03ff-44ad-8d5c-42bd496c3100","private_publication":false,"abstract":"Pose Machines provide a sequential prediction frame-work for learning rich implicit spatial models. In this work we show a systematic design for how convolutional net-works can be incorporated into the pose machine frame-work for learning image features and image-dependent spa-tial models for the task of pose estimation. The contribution of this paper is to implicitly model long-range dependen-cies between variables in structured prediction tasks such as articulated pose estimation. We achieve this by designing a sequential architecture composed of convolutional net-works that directly operate on belief maps from previous stages, producing increasingly refined estimates for part lo-cations, without the need for explicit graphical model-style inference. Our approach addresses the characteristic diffi-culty of vanishing gradients during training by providing a natural learning objective function that enforces intermedi-ate supervision, thereby replenishing back-propagated gra-dients and conditioning the learning procedure. We demon-strate state-of-the-art performance and outperform compet-ing methods on standard benchmarks including the MPII, LSP, and FLIC datasets.","bibtype":"article","author":"Wei, Shih-En and Ramakrishna, Varun and Kanada, Takeo and Sheikh, Yaser","journal":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition","bibtex":"@article{\n title = {Pose Machines :},\n type = {article},\n year = {2016},\n id = {5ccfc8d8-0bf4-334c-a755-dc378ba6f7c0},\n created = {2022-09-19T10:49:11.819Z},\n file_attached = {true},\n profile_id = {276016a7-2c9d-3507-8888-093db7c54774},\n group_id = {5ec9cc91-a5d6-3de5-82f3-3ef3d98a89c1},\n last_modified = {2022-09-19T10:50:34.378Z},\n read = {false},\n starred = {false},\n authored = {false},\n confirmed = {true},\n hidden = {false},\n folder_uuids = {02fb5526-03ff-44ad-8d5c-42bd496c3100},\n private_publication = {false},\n abstract = {Pose Machines provide a sequential prediction frame-work for learning rich implicit spatial models. In this work we show a systematic design for how convolutional net-works can be incorporated into the pose machine frame-work for learning image features and image-dependent spa-tial models for the task of pose estimation. The contribution of this paper is to implicitly model long-range dependen-cies between variables in structured prediction tasks such as articulated pose estimation. We achieve this by designing a sequential architecture composed of convolutional net-works that directly operate on belief maps from previous stages, producing increasingly refined estimates for part lo-cations, without the need for explicit graphical model-style inference. Our approach addresses the characteristic diffi-culty of vanishing gradients during training by providing a natural learning objective function that enforces intermedi-ate supervision, thereby replenishing back-propagated gra-dients and conditioning the learning procedure. We demon-strate state-of-the-art performance and outperform compet-ing methods on standard benchmarks including the MPII, LSP, and FLIC datasets.},\n bibtype = {article},\n author = {Wei, Shih-En and Ramakrishna, Varun and Kanada, Takeo and Sheikh, Yaser},\n journal = {Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition}\n}","author_short":["Wei, S.","Ramakrishna, V.","Kanada, T.","Sheikh, Y."],"urls":{"Paper":"https://bibbase.org/service/mendeley/bfbbf840-4c42-3914-a463-19024f50b30c/file/3485ebbe-8256-22a7-430a-361e263e0da2/Wei_Convolutional_Pose_Machines_CVPR_2016_paper.pdf.pdf"},"biburl":"https://bibbase.org/service/mendeley/bfbbf840-4c42-3914-a463-19024f50b30c","bibbaseid":"wei-ramakrishna-kanada-sheikh-posemachines-2016","role":"author","metadata":{"authorlinks":{}},"downloads":0},"bibtype":"article","biburl":"https://bibbase.org/service/mendeley/bfbbf840-4c42-3914-a463-19024f50b30c","dataSources":["2252seNhipfTmjEBQ"],"keywords":[],"search_terms":["pose","machines","wei","ramakrishna","kanada","sheikh"],"title":"Pose Machines :","year":2016}