Photo Filter Classification and Filter Recommendation without Much Manual Labeling

Wei Ta Chu; Yu Tzu Fan

doi:10.1109/MMSP.2019.8901831

Photo Filter Classification and Filter Recommendation without Much Manual Labeling

Wei Ta Chu, Yu Tzu Fan

Department of Computer Science and Information Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

Because how users employ filters to photos may reveal user's preference or mental state, a photo filter classification method is potentially demanded to enable future large-scale analysis. We adopt the transfer learning technique to transform deep models pre-trained for object classification into models suitable for photo filter classification. Based on accurate classification results, we build a filter recommendation approach without much manual labeling. It can be easily extended when more training data are available. A series of experimental studies are conducted to demonstrate effectiveness of filter classification with transfer learning. We also demonstrate the proposed filter recommendation achieves encouraging performance.

Original language	English
Title of host publication	IEEE 21st International Workshop on Multimedia Signal Processing, MMSP 2019
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9781728118178
DOIs	https://doi.org/10.1109/MMSP.2019.8901831
Publication status	Published - 2019 Sep
Event	21st IEEE International Workshop on Multimedia Signal Processing, MMSP 2019 - Kuala Lumpur, Malaysia Duration: 2019 Sep 27 → 2019 Sep 29

Publication series

Name	IEEE 21st International Workshop on Multimedia Signal Processing, MMSP 2019

Conference

Conference	21st IEEE International Workshop on Multimedia Signal Processing, MMSP 2019
Country	Malaysia
City	Kuala Lumpur
Period	19-09-27 → 19-09-29

All Science Journal Classification (ASJC) codes

Signal Processing
Media Technology

Access to Document

10.1109/MMSP.2019.8901831

Fingerprint Dive into the research topics of 'Photo Filter Classification and Filter Recommendation without Much Manual Labeling'. Together they form a unique fingerprint.

Cite this

Chu, W. T., & Fan, Y. T. (2019). Photo Filter Classification and Filter Recommendation without Much Manual Labeling. In IEEE 21st International Workshop on Multimedia Signal Processing, MMSP 2019 [8901831] (IEEE 21st International Workshop on Multimedia Signal Processing, MMSP 2019). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/MMSP.2019.8901831

@inproceedings{6e1c520eac854ede9ed04fccbfe58669,

title = "Photo Filter Classification and Filter Recommendation without Much Manual Labeling",

abstract = "Because how users employ filters to photos may reveal user's preference or mental state, a photo filter classification method is potentially demanded to enable future large-scale analysis. We adopt the transfer learning technique to transform deep models pre-trained for object classification into models suitable for photo filter classification. Based on accurate classification results, we build a filter recommendation approach without much manual labeling. It can be easily extended when more training data are available. A series of experimental studies are conducted to demonstrate effectiveness of filter classification with transfer learning. We also demonstrate the proposed filter recommendation achieves encouraging performance.",

author = "Chu, {Wei Ta} and Fan, {Yu Tzu}",

year = "2019",

month = sep,

doi = "10.1109/MMSP.2019.8901831",

language = "English",

series = "IEEE 21st International Workshop on Multimedia Signal Processing, MMSP 2019",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

booktitle = "IEEE 21st International Workshop on Multimedia Signal Processing, MMSP 2019",

address = "United States",

note = "21st IEEE International Workshop on Multimedia Signal Processing, MMSP 2019 ; Conference date: 27-09-2019 Through 29-09-2019",

}

Chu, WT & Fan, YT 2019, Photo Filter Classification and Filter Recommendation without Much Manual Labeling. in IEEE 21st International Workshop on Multimedia Signal Processing, MMSP 2019., 8901831, IEEE 21st International Workshop on Multimedia Signal Processing, MMSP 2019, Institute of Electrical and Electronics Engineers Inc., 21st IEEE International Workshop on Multimedia Signal Processing, MMSP 2019, Kuala Lumpur, Malaysia, 19-09-27. https://doi.org/10.1109/MMSP.2019.8901831

Photo Filter Classification and Filter Recommendation without Much Manual Labeling. / Chu, Wei Ta; Fan, Yu Tzu.

IEEE 21st International Workshop on Multimedia Signal Processing, MMSP 2019. Institute of Electrical and Electronics Engineers Inc., 2019. 8901831 (IEEE 21st International Workshop on Multimedia Signal Processing, MMSP 2019).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Photo Filter Classification and Filter Recommendation without Much Manual Labeling

AU - Chu, Wei Ta

AU - Fan, Yu Tzu

PY - 2019/9

Y1 - 2019/9

N2 - Because how users employ filters to photos may reveal user's preference or mental state, a photo filter classification method is potentially demanded to enable future large-scale analysis. We adopt the transfer learning technique to transform deep models pre-trained for object classification into models suitable for photo filter classification. Based on accurate classification results, we build a filter recommendation approach without much manual labeling. It can be easily extended when more training data are available. A series of experimental studies are conducted to demonstrate effectiveness of filter classification with transfer learning. We also demonstrate the proposed filter recommendation achieves encouraging performance.

AB - Because how users employ filters to photos may reveal user's preference or mental state, a photo filter classification method is potentially demanded to enable future large-scale analysis. We adopt the transfer learning technique to transform deep models pre-trained for object classification into models suitable for photo filter classification. Based on accurate classification results, we build a filter recommendation approach without much manual labeling. It can be easily extended when more training data are available. A series of experimental studies are conducted to demonstrate effectiveness of filter classification with transfer learning. We also demonstrate the proposed filter recommendation achieves encouraging performance.

UR - http://www.scopus.com/inward/record.url?scp=85075757150&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85075757150&partnerID=8YFLogxK

U2 - 10.1109/MMSP.2019.8901831

DO - 10.1109/MMSP.2019.8901831

M3 - Conference contribution

T3 - IEEE 21st International Workshop on Multimedia Signal Processing, MMSP 2019

BT - IEEE 21st International Workshop on Multimedia Signal Processing, MMSP 2019

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 21st IEEE International Workshop on Multimedia Signal Processing, MMSP 2019

Y2 - 27 September 2019 through 29 September 2019

ER -

Chu WT, Fan YT. Photo Filter Classification and Filter Recommendation without Much Manual Labeling. In IEEE 21st International Workshop on Multimedia Signal Processing, MMSP 2019. Institute of Electrical and Electronics Engineers Inc. 2019. 8901831. (IEEE 21st International Workshop on Multimedia Signal Processing, MMSP 2019). https://doi.org/10.1109/MMSP.2019.8901831