Estimation of Missing Human Body Parts via Bidirectional LSTM

Ibrahim Hamed Ismail RADWAN, Akshay Asthana, Hafsa ISMAIL, Byron Keating, Roland GOECKE

Research output: A Conference proceeding or a Chapter in BookConference contribution

Abstract

In this paper, a bi-directional long-short term memory (LSTM) based approach is proposed for the estimation of missing body parts in a human pose estimation context. Accurate human pose estimation is often a key component for accurate human action and activity recognition. The key idea of our algorithm is to learn the temporal consistencies of the human body poses between previous and subsequent frames. This helps in estimating missing body parts and improves the general smoothness of the pose detection results. The approach acts as a post-processing step after the application of any off-the-shelf body part detector and has been evaluated on the PoseTrack dataset for both validation and testing sequences. The results show consistent improvement in the detection across all body parts.
Original languageEnglish
Title of host publicationProceedings of the 2019 14th IEEE International Conference on Automatic Face & Gesture Recognition (FG 2019)
Place of PublicationDanvers, USA
PublisherIEEE, Institute of Electrical and Electronics Engineers
Pages443-447
Number of pages5
ISBN (Print)9781728100890
DOIs
Publication statusPublished - 2019
Event14th IEEE International Conference on Automatic Face and Gesture Recognition - Lille, France
Duration: 14 May 201918 May 2019
http://fg2019.org/

Conference

Conference14th IEEE International Conference on Automatic Face and Gesture Recognition
Abbreviated titleFG 2019
CountryFrance
CityLille
Period14/05/1918/05/19
Internet address

Fingerprint

Detectors
Testing
Processing
Long short-term memory

Cite this

RADWAN, I. H. I., Asthana, A., ISMAIL, H., Keating, B., & GOECKE, R. (2019). Estimation of Missing Human Body Parts via Bidirectional LSTM. In Proceedings of the 2019 14th IEEE International Conference on Automatic Face & Gesture Recognition (FG 2019) (pp. 443-447). Danvers, USA: IEEE, Institute of Electrical and Electronics Engineers. https://doi.org/10.1109/FG.2019.8756597
RADWAN, Ibrahim Hamed Ismail ; Asthana, Akshay ; ISMAIL, Hafsa ; Keating, Byron ; GOECKE, Roland. / Estimation of Missing Human Body Parts via Bidirectional LSTM. Proceedings of the 2019 14th IEEE International Conference on Automatic Face & Gesture Recognition (FG 2019). Danvers, USA : IEEE, Institute of Electrical and Electronics Engineers, 2019. pp. 443-447
@inproceedings{9294e4cfd85d43919fde1f374430590b,
title = "Estimation of Missing Human Body Parts via Bidirectional LSTM",
abstract = "In this paper, a bi-directional long-short term memory (LSTM) based approach is proposed for the estimation of missing body parts in a human pose estimation context. Accurate human pose estimation is often a key component for accurate human action and activity recognition. The key idea of our algorithm is to learn the temporal consistencies of the human body poses between previous and subsequent frames. This helps in estimating missing body parts and improves the general smoothness of the pose detection results. The approach acts as a post-processing step after the application of any off-the-shelf body part detector and has been evaluated on the PoseTrack dataset for both validation and testing sequences. The results show consistent improvement in the detection across all body parts.",
keywords = "Human pose estimation, Bidirectional LSTM",
author = "RADWAN, {Ibrahim Hamed Ismail} and Akshay Asthana and Hafsa ISMAIL and Byron Keating and Roland GOECKE",
year = "2019",
doi = "10.1109/FG.2019.8756597",
language = "English",
isbn = "9781728100890",
pages = "443--447",
booktitle = "Proceedings of the 2019 14th IEEE International Conference on Automatic Face & Gesture Recognition (FG 2019)",
publisher = "IEEE, Institute of Electrical and Electronics Engineers",
address = "United States",

}

RADWAN, IHI, Asthana, A, ISMAIL, H, Keating, B & GOECKE, R 2019, Estimation of Missing Human Body Parts via Bidirectional LSTM. in Proceedings of the 2019 14th IEEE International Conference on Automatic Face & Gesture Recognition (FG 2019). IEEE, Institute of Electrical and Electronics Engineers, Danvers, USA, pp. 443-447, 14th IEEE International Conference on Automatic Face and Gesture Recognition, Lille, France, 14/05/19. https://doi.org/10.1109/FG.2019.8756597

Estimation of Missing Human Body Parts via Bidirectional LSTM. / RADWAN, Ibrahim Hamed Ismail; Asthana, Akshay; ISMAIL, Hafsa; Keating, Byron; GOECKE, Roland.

Proceedings of the 2019 14th IEEE International Conference on Automatic Face & Gesture Recognition (FG 2019). Danvers, USA : IEEE, Institute of Electrical and Electronics Engineers, 2019. p. 443-447.

Research output: A Conference proceeding or a Chapter in BookConference contribution

TY - GEN

T1 - Estimation of Missing Human Body Parts via Bidirectional LSTM

AU - RADWAN, Ibrahim Hamed Ismail

AU - Asthana, Akshay

AU - ISMAIL, Hafsa

AU - Keating, Byron

AU - GOECKE, Roland

PY - 2019

Y1 - 2019

N2 - In this paper, a bi-directional long-short term memory (LSTM) based approach is proposed for the estimation of missing body parts in a human pose estimation context. Accurate human pose estimation is often a key component for accurate human action and activity recognition. The key idea of our algorithm is to learn the temporal consistencies of the human body poses between previous and subsequent frames. This helps in estimating missing body parts and improves the general smoothness of the pose detection results. The approach acts as a post-processing step after the application of any off-the-shelf body part detector and has been evaluated on the PoseTrack dataset for both validation and testing sequences. The results show consistent improvement in the detection across all body parts.

AB - In this paper, a bi-directional long-short term memory (LSTM) based approach is proposed for the estimation of missing body parts in a human pose estimation context. Accurate human pose estimation is often a key component for accurate human action and activity recognition. The key idea of our algorithm is to learn the temporal consistencies of the human body poses between previous and subsequent frames. This helps in estimating missing body parts and improves the general smoothness of the pose detection results. The approach acts as a post-processing step after the application of any off-the-shelf body part detector and has been evaluated on the PoseTrack dataset for both validation and testing sequences. The results show consistent improvement in the detection across all body parts.

KW - Human pose estimation

KW - Bidirectional LSTM

UR - http://www.scopus.com/inward/record.url?scp=85070438798&partnerID=8YFLogxK

UR - http://www.mendeley.com/research/estimation-missing-human-body-parts-via-bidirectional-lstm

U2 - 10.1109/FG.2019.8756597

DO - 10.1109/FG.2019.8756597

M3 - Conference contribution

SN - 9781728100890

SP - 443

EP - 447

BT - Proceedings of the 2019 14th IEEE International Conference on Automatic Face & Gesture Recognition (FG 2019)

PB - IEEE, Institute of Electrical and Electronics Engineers

CY - Danvers, USA

ER -

RADWAN IHI, Asthana A, ISMAIL H, Keating B, GOECKE R. Estimation of Missing Human Body Parts via Bidirectional LSTM. In Proceedings of the 2019 14th IEEE International Conference on Automatic Face & Gesture Recognition (FG 2019). Danvers, USA: IEEE, Institute of Electrical and Electronics Engineers. 2019. p. 443-447 https://doi.org/10.1109/FG.2019.8756597