Semi-Supervised Classification With Pairwise Constraints: A Case Study on Animal Identification from Video

Allbwn ymchwil: Cyfraniad at gyfnodolynErthygladolygiad gan gymheiriaid

StandardStandard

Semi-Supervised Classification With Pairwise Constraints: A Case Study on Animal Identification from Video. / Kuncheva, Ludmila; Garrido-Labrador, Jose; Ramos-Perez, Ismael et al.
Yn: Information Fusion, Cyfrol 104, 102188, 01.04.2024.

Allbwn ymchwil: Cyfraniad at gyfnodolynErthygladolygiad gan gymheiriaid

HarvardHarvard

Kuncheva, L, Garrido-Labrador, J, Ramos-Perez, I, Hennessey, S & Rodriguez, J 2024, 'Semi-Supervised Classification With Pairwise Constraints: A Case Study on Animal Identification from Video', Information Fusion, cyfrol. 104, 102188. https://doi.org/10.1016/j.inffus.2023.102188

APA

Kuncheva, L., Garrido-Labrador, J., Ramos-Perez, I., Hennessey, S., & Rodriguez, J. (2024). Semi-Supervised Classification With Pairwise Constraints: A Case Study on Animal Identification from Video. Information Fusion, 104, Erthygl 102188. https://doi.org/10.1016/j.inffus.2023.102188

CBE

Kuncheva L, Garrido-Labrador J, Ramos-Perez I, Hennessey S, Rodriguez J. 2024. Semi-Supervised Classification With Pairwise Constraints: A Case Study on Animal Identification from Video. Information Fusion. 104:Article 102188. https://doi.org/10.1016/j.inffus.2023.102188

MLA

VancouverVancouver

Kuncheva L, Garrido-Labrador J, Ramos-Perez I, Hennessey S, Rodriguez J. Semi-Supervised Classification With Pairwise Constraints: A Case Study on Animal Identification from Video. Information Fusion. 2024 Ebr 1;104:102188. Epub 2023 Rhag 12. doi: 10.1016/j.inffus.2023.102188

Author

Kuncheva, Ludmila ; Garrido-Labrador, Jose ; Ramos-Perez, Ismael et al. / Semi-Supervised Classification With Pairwise Constraints: A Case Study on Animal Identification from Video. Yn: Information Fusion. 2024 ; Cyfrol 104.

RIS

TY - JOUR

T1 - Semi-Supervised Classification With Pairwise Constraints: A Case Study on Animal Identification from Video

AU - Kuncheva, Ludmila

AU - Garrido-Labrador, Jose

AU - Ramos-Perez, Ismael

AU - Hennessey, Samuel

AU - Rodriguez, Juan

PY - 2024/4/1

Y1 - 2024/4/1

N2 - Mainstream semi-supervised classification assumes that part of the available data are labelled. Here we assume that, in addition to the labels, we have pairwise constraints on the unlabelled data. Each constraint links two instances, and is one of Must Link (ML, belong to the same class) or Cannot Link (CL, belong to different classes). We propose an approach that uses the labelled data to train a classifier and then applies the ML and CL constraints in subsequent labelling. In our approach, a set of instances are labelled at the same time. Our case study is on animal re-identification. The dataset consists of five free-camera video clips of of animals (koi fish, pigeons and pigs), annotated with bounding boxes and animal identities. The proposed approach combines the representations or classifiers predictions from the bounding boxes of consecutive frames. We demonstrate that our approach outperforms standard classifiers, constrained clustering, as well as inductive and transductive semi-supervised learning, using five feature representations.

AB - Mainstream semi-supervised classification assumes that part of the available data are labelled. Here we assume that, in addition to the labels, we have pairwise constraints on the unlabelled data. Each constraint links two instances, and is one of Must Link (ML, belong to the same class) or Cannot Link (CL, belong to different classes). We propose an approach that uses the labelled data to train a classifier and then applies the ML and CL constraints in subsequent labelling. In our approach, a set of instances are labelled at the same time. Our case study is on animal re-identification. The dataset consists of five free-camera video clips of of animals (koi fish, pigeons and pigs), annotated with bounding boxes and animal identities. The proposed approach combines the representations or classifiers predictions from the bounding boxes of consecutive frames. We demonstrate that our approach outperforms standard classifiers, constrained clustering, as well as inductive and transductive semi-supervised learning, using five feature representations.

U2 - 10.1016/j.inffus.2023.102188

DO - 10.1016/j.inffus.2023.102188

M3 - Article

VL - 104

JO - Information Fusion

JF - Information Fusion

SN - 1566-2535

M1 - 102188

ER -