Stacked Hourglass Networks for Human Pose Estimation
A. Newell, K. Yang, and J. Deng. Computer Vision -- ECCV 2016, page 483--499. Cham, Springer International Publishing, (2016)
Abstract
This work introduces a novel convolutional network architecture for the task of human pose estimation. Features are processed across all scales and consolidated to best capture the various spatial relationships associated with the body. We show how repeated bottom-up, top-down processing used in conjunction with intermediate supervision is critical to improving the performance of the network. We refer to the architecture as a ``stacked hourglass'' network based on the successive steps of pooling and upsampling that are done to produce a final set of predictions. State-of-the-art results are achieved on the FLIC and MPII benchmarks outcompeting all recent methods.
Description
Stacked Hourglass Networks for Human Pose Estimation | SpringerLink
%0 Conference Paper
%1 10.1007/978-3-319-46484-8_29
%A Newell, Alejandro
%A Yang, Kaiyu
%A Deng, Jia
%B Computer Vision -- ECCV 2016
%C Cham
%D 2016
%E Leibe, Bastian
%E Matas, Jiri
%E Sebe, Nicu
%E Welling, Max
%I Springer International Publishing
%K convnets dnn pose
%P 483--499
%T Stacked Hourglass Networks for Human Pose Estimation
%X This work introduces a novel convolutional network architecture for the task of human pose estimation. Features are processed across all scales and consolidated to best capture the various spatial relationships associated with the body. We show how repeated bottom-up, top-down processing used in conjunction with intermediate supervision is critical to improving the performance of the network. We refer to the architecture as a ``stacked hourglass'' network based on the successive steps of pooling and upsampling that are done to produce a final set of predictions. State-of-the-art results are achieved on the FLIC and MPII benchmarks outcompeting all recent methods.
%@ 978-3-319-46484-8
@inproceedings{10.1007/978-3-319-46484-8_29,
abstract = {This work introduces a novel convolutional network architecture for the task of human pose estimation. Features are processed across all scales and consolidated to best capture the various spatial relationships associated with the body. We show how repeated bottom-up, top-down processing used in conjunction with intermediate supervision is critical to improving the performance of the network. We refer to the architecture as a ``stacked hourglass'' network based on the successive steps of pooling and upsampling that are done to produce a final set of predictions. State-of-the-art results are achieved on the FLIC and MPII benchmarks outcompeting all recent methods.},
added-at = {2020-10-16T18:49:55.000+0200},
address = {Cham},
author = {Newell, Alejandro and Yang, Kaiyu and Deng, Jia},
biburl = {https://www.bibsonomy.org/bibtex/2e301d2451d77cc685196c989bca4732b/sohnki},
booktitle = {Computer Vision -- ECCV 2016},
description = {Stacked Hourglass Networks for Human Pose Estimation | SpringerLink},
editor = {Leibe, Bastian and Matas, Jiri and Sebe, Nicu and Welling, Max},
interhash = {1943d16096c0d502d34e13ccbdf5d82f},
intrahash = {e301d2451d77cc685196c989bca4732b},
isbn = {978-3-319-46484-8},
keywords = {convnets dnn pose},
pages = {483--499},
publisher = {Springer International Publishing},
timestamp = {2020-10-16T18:49:55.000+0200},
title = {Stacked Hourglass Networks for Human Pose Estimation},
year = 2016
}