Beyond IID: Learning to combine Non-IID metrics for vision tasks

Yinghuan Shi, Wenbin Li, Yang Gao, Longbing Cao, Dinggang Shen

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Citations (Scopus)

Abstract

Metric learning has been widely employed, especially in various computer vision tasks, with the fundamental assumption that all samples (e.g., regions/superpixels in images/videos) are independent and identically distributed (IID). However, since the samples are usually spatially-connected or temporally-correlated with their physically-connected neighbours, they are not IID (non-IID for short), which cannot be directly handled by existing methods. Thus, we propose to learn and integrate non-IID metrics (NIME). To incorporate the non-IID spatial/temporal relations, instead of directly using non-IID features and metric learning as previous methods, NIME first builds several non-IID representations on original (non-IID) features by various graph kernel functions, and then automatically learns the metric under the best combination of various non-IID representations. NIME is applied to solve two typical computer vision tasks: interactive image segmentation and histology image identification. The results show that learning and integrating non-IID metrics improves the performance, compared to the IID methods. Moreover, our method achieves results comparable or better than that of the state-of-the-arts.

Original languageEnglish
Title of host publication31st AAAI Conference on Artificial Intelligence, AAAI 2017
PublisherAAAI press
Pages1524-1531
Number of pages8
Publication statusPublished - 2017 Jan 1
Externally publishedYes
Event31st AAAI Conference on Artificial Intelligence, AAAI 2017 - San Francisco, United States
Duration: 2017 Feb 42017 Feb 10

Other

Other31st AAAI Conference on Artificial Intelligence, AAAI 2017
CountryUnited States
CitySan Francisco
Period17/2/417/2/10

Fingerprint

Computer vision
Histology
Image segmentation

ASJC Scopus subject areas

  • Artificial Intelligence

Cite this

Shi, Y., Li, W., Gao, Y., Cao, L., & Shen, D. (2017). Beyond IID: Learning to combine Non-IID metrics for vision tasks. In 31st AAAI Conference on Artificial Intelligence, AAAI 2017 (pp. 1524-1531). AAAI press.

Beyond IID : Learning to combine Non-IID metrics for vision tasks. / Shi, Yinghuan; Li, Wenbin; Gao, Yang; Cao, Longbing; Shen, Dinggang.

31st AAAI Conference on Artificial Intelligence, AAAI 2017. AAAI press, 2017. p. 1524-1531.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Shi, Y, Li, W, Gao, Y, Cao, L & Shen, D 2017, Beyond IID: Learning to combine Non-IID metrics for vision tasks. in 31st AAAI Conference on Artificial Intelligence, AAAI 2017. AAAI press, pp. 1524-1531, 31st AAAI Conference on Artificial Intelligence, AAAI 2017, San Francisco, United States, 17/2/4.
Shi Y, Li W, Gao Y, Cao L, Shen D. Beyond IID: Learning to combine Non-IID metrics for vision tasks. In 31st AAAI Conference on Artificial Intelligence, AAAI 2017. AAAI press. 2017. p. 1524-1531
Shi, Yinghuan ; Li, Wenbin ; Gao, Yang ; Cao, Longbing ; Shen, Dinggang. / Beyond IID : Learning to combine Non-IID metrics for vision tasks. 31st AAAI Conference on Artificial Intelligence, AAAI 2017. AAAI press, 2017. pp. 1524-1531
@inproceedings{4bfc3fe3e4d54e209e46f4da4dcc9c28,
title = "Beyond IID: Learning to combine Non-IID metrics for vision tasks",
abstract = "Metric learning has been widely employed, especially in various computer vision tasks, with the fundamental assumption that all samples (e.g., regions/superpixels in images/videos) are independent and identically distributed (IID). However, since the samples are usually spatially-connected or temporally-correlated with their physically-connected neighbours, they are not IID (non-IID for short), which cannot be directly handled by existing methods. Thus, we propose to learn and integrate non-IID metrics (NIME). To incorporate the non-IID spatial/temporal relations, instead of directly using non-IID features and metric learning as previous methods, NIME first builds several non-IID representations on original (non-IID) features by various graph kernel functions, and then automatically learns the metric under the best combination of various non-IID representations. NIME is applied to solve two typical computer vision tasks: interactive image segmentation and histology image identification. The results show that learning and integrating non-IID metrics improves the performance, compared to the IID methods. Moreover, our method achieves results comparable or better than that of the state-of-the-arts.",
author = "Yinghuan Shi and Wenbin Li and Yang Gao and Longbing Cao and Dinggang Shen",
year = "2017",
month = "1",
day = "1",
language = "English",
pages = "1524--1531",
booktitle = "31st AAAI Conference on Artificial Intelligence, AAAI 2017",
publisher = "AAAI press",

}

TY - GEN

T1 - Beyond IID

T2 - Learning to combine Non-IID metrics for vision tasks

AU - Shi, Yinghuan

AU - Li, Wenbin

AU - Gao, Yang

AU - Cao, Longbing

AU - Shen, Dinggang

PY - 2017/1/1

Y1 - 2017/1/1

N2 - Metric learning has been widely employed, especially in various computer vision tasks, with the fundamental assumption that all samples (e.g., regions/superpixels in images/videos) are independent and identically distributed (IID). However, since the samples are usually spatially-connected or temporally-correlated with their physically-connected neighbours, they are not IID (non-IID for short), which cannot be directly handled by existing methods. Thus, we propose to learn and integrate non-IID metrics (NIME). To incorporate the non-IID spatial/temporal relations, instead of directly using non-IID features and metric learning as previous methods, NIME first builds several non-IID representations on original (non-IID) features by various graph kernel functions, and then automatically learns the metric under the best combination of various non-IID representations. NIME is applied to solve two typical computer vision tasks: interactive image segmentation and histology image identification. The results show that learning and integrating non-IID metrics improves the performance, compared to the IID methods. Moreover, our method achieves results comparable or better than that of the state-of-the-arts.

AB - Metric learning has been widely employed, especially in various computer vision tasks, with the fundamental assumption that all samples (e.g., regions/superpixels in images/videos) are independent and identically distributed (IID). However, since the samples are usually spatially-connected or temporally-correlated with their physically-connected neighbours, they are not IID (non-IID for short), which cannot be directly handled by existing methods. Thus, we propose to learn and integrate non-IID metrics (NIME). To incorporate the non-IID spatial/temporal relations, instead of directly using non-IID features and metric learning as previous methods, NIME first builds several non-IID representations on original (non-IID) features by various graph kernel functions, and then automatically learns the metric under the best combination of various non-IID representations. NIME is applied to solve two typical computer vision tasks: interactive image segmentation and histology image identification. The results show that learning and integrating non-IID metrics improves the performance, compared to the IID methods. Moreover, our method achieves results comparable or better than that of the state-of-the-arts.

UR - http://www.scopus.com/inward/record.url?scp=85030479431&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85030479431&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:85030479431

SP - 1524

EP - 1531

BT - 31st AAAI Conference on Artificial Intelligence, AAAI 2017

PB - AAAI press

ER -