Classification and Weakly Supervised Pain Localization using Multiple Segment Representation

Karan Sikka, Abhinav Dhall, Marion Bartlett

Research output: Contribution to journalArticle

23 Citations (Scopus)

Abstract

Automatic pain recognition from videos is a vital clinical application and, owing to its spontaneous nature, poses interesting challenges to automatic facial expression recognition (AFER) research. Previous pain vs no-pain systems have highlighted two major challenges: (1) ground truth is provided for the sequence, but the presence or absence of the target expression for a given frame is unknown, and (2) the time point and the duration of the pain expression event(s) in each video are unknown. To address these issues we propose a novel framework (referred to as MS-MIL) where each sequence is represented as a bag containing multiple segments, and multiple instance learning (MIL) is employed to handle this weakly labeled data in the form of sequence level ground-truth.
Original languageEnglish
Pages (from-to)659-670
Number of pages12
JournalImage and Vision Computing
Volume32
Issue number10
DOIs
Publication statusPublished - 2014
Externally publishedYes

Cite this

Sikka, Karan ; Dhall, Abhinav ; Bartlett, Marion. / Classification and Weakly Supervised Pain Localization using Multiple Segment Representation. In: Image and Vision Computing. 2014 ; Vol. 32, No. 10. pp. 659-670.
@article{6d09f6f8a46f43229d5e485f79d36050,
title = "Classification and Weakly Supervised Pain Localization using Multiple Segment Representation",
abstract = "Automatic pain recognition from videos is a vital clinical application and, owing to its spontaneous nature, poses interesting challenges to automatic facial expression recognition (AFER) research. Previous pain vs no-pain systems have highlighted two major challenges: (1) ground truth is provided for the sequence, but the presence or absence of the target expression for a given frame is unknown, and (2) the time point and the duration of the pain expression event(s) in each video are unknown. To address these issues we propose a novel framework (referred to as MS-MIL) where each sequence is represented as a bag containing multiple segments, and multiple instance learning (MIL) is employed to handle this weakly labeled data in the form of sequence level ground-truth.",
keywords = "Action classification, Bag of Words, Bagging, Boosting, Emotion classification, Pain, Temporal Segmentation, Weakly Supervised Learning",
author = "Karan Sikka and Abhinav Dhall and Marion Bartlett",
year = "2014",
doi = "10.1016/j.imavis.2014.02.008",
language = "English",
volume = "32",
pages = "659--670",
journal = "Image and Vision Computing",
issn = "0262-8856",
publisher = "Elsevier Limited",
number = "10",

}

Classification and Weakly Supervised Pain Localization using Multiple Segment Representation. / Sikka, Karan; Dhall, Abhinav; Bartlett, Marion.

In: Image and Vision Computing, Vol. 32, No. 10, 2014, p. 659-670.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Classification and Weakly Supervised Pain Localization using Multiple Segment Representation

AU - Sikka, Karan

AU - Dhall, Abhinav

AU - Bartlett, Marion

PY - 2014

Y1 - 2014

N2 - Automatic pain recognition from videos is a vital clinical application and, owing to its spontaneous nature, poses interesting challenges to automatic facial expression recognition (AFER) research. Previous pain vs no-pain systems have highlighted two major challenges: (1) ground truth is provided for the sequence, but the presence or absence of the target expression for a given frame is unknown, and (2) the time point and the duration of the pain expression event(s) in each video are unknown. To address these issues we propose a novel framework (referred to as MS-MIL) where each sequence is represented as a bag containing multiple segments, and multiple instance learning (MIL) is employed to handle this weakly labeled data in the form of sequence level ground-truth.

AB - Automatic pain recognition from videos is a vital clinical application and, owing to its spontaneous nature, poses interesting challenges to automatic facial expression recognition (AFER) research. Previous pain vs no-pain systems have highlighted two major challenges: (1) ground truth is provided for the sequence, but the presence or absence of the target expression for a given frame is unknown, and (2) the time point and the duration of the pain expression event(s) in each video are unknown. To address these issues we propose a novel framework (referred to as MS-MIL) where each sequence is represented as a bag containing multiple segments, and multiple instance learning (MIL) is employed to handle this weakly labeled data in the form of sequence level ground-truth.

KW - Action classification

KW - Bag of Words

KW - Bagging

KW - Boosting

KW - Emotion classification

KW - Pain

KW - Temporal Segmentation

KW - Weakly Supervised Learning

U2 - 10.1016/j.imavis.2014.02.008

DO - 10.1016/j.imavis.2014.02.008

M3 - Article

VL - 32

SP - 659

EP - 670

JO - Image and Vision Computing

JF - Image and Vision Computing

SN - 0262-8856

IS - 10

ER -