A New Framework for Evaluating Image Quality Including Deep Learning Task Performances as a Proxy
<sc>iquaflow</sc> is a framework that provides a set of tools to assess image quality. The user can add custom metrics that can be easily integrated and a set of unsupervised methods is offered by default. Furthermore, <sc>iquaflow</sc> measures quality by using the performan...
| Main Authors: | , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2024-01-01
|
| Series: | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10356628/ |
| _version_ | 1797330382105870336 |
|---|---|
| author | Pau Galles Katalin Takats Miguel Hernandez-Cabronero David Berga Luciano Pega Laura Riordan-Chen Clara Garcia Guillermo Becker Adan Garriga Anica Bukva Joan Serra-Sagrista David Vilaseca Javier Marin |
| author_facet | Pau Galles Katalin Takats Miguel Hernandez-Cabronero David Berga Luciano Pega Laura Riordan-Chen Clara Garcia Guillermo Becker Adan Garriga Anica Bukva Joan Serra-Sagrista David Vilaseca Javier Marin |
| author_sort | Pau Galles |
| collection | DOAJ |
| description | <sc>iquaflow</sc> is a framework that provides a set of tools to assess image quality. The user can add custom metrics that can be easily integrated and a set of unsupervised methods is offered by default. Furthermore, <sc>iquaflow</sc> measures quality by using the performance of AI models trained on the images as a proxy. This also helps to easily make studies of performance degradation of several modifications of the original dataset, for instance, with images reconstructed after different levels of lossy compression; satellite images would be a use case example, since they are commonly compressed before downloading to the ground. In this situation, the optimization problem involves finding images that, while being compressed to their smallest possible file size, still maintain sufficient quality to meet the required performance of the deep learning algorithms. Thus, a study with <sc>iquaflow</sc> is suitable for such case. All this development is wrapped in <sc>Mlflow</sc>: an interactive tool used to visualize and summarize the results. This document describes different use cases and provides links to their respective repositories. To ease the creation of new studies, we include a cookiecutter repository. The source code, issue tracker and aforementioned repositories are all hosted on GitHub. |
| first_indexed | 2024-03-08T07:19:04Z |
| format | Article |
| id | doaj.art-18c255c09b0a477aae1dc4e4b8afad95 |
| institution | Directory Open Access Journal |
| issn | 2151-1535 |
| language | English |
| last_indexed | 2024-03-08T07:19:04Z |
| publishDate | 2024-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| spelling | doaj.art-18c255c09b0a477aae1dc4e4b8afad952024-02-03T00:01:57ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing2151-15352024-01-01173285329610.1109/JSTARS.2023.334247510356628A New Framework for Evaluating Image Quality Including Deep Learning Task Performances as a ProxyPau Galles0https://orcid.org/0000-0002-1893-9353Katalin Takats1https://orcid.org/0000-0001-6636-3195Miguel Hernandez-Cabronero2https://orcid.org/0000-0001-9301-4337David Berga3https://orcid.org/0000-0001-7543-2770Luciano Pega4https://orcid.org/0009-0003-6878-8156Laura Riordan-Chen5https://orcid.org/0009-0003-9883-8505Clara Garcia6https://orcid.org/0000-0002-1662-1747Guillermo Becker7https://orcid.org/0009-0003-6114-9909Adan Garriga8https://orcid.org/0000-0003-2681-4005Anica Bukva9https://orcid.org/0009-0002-6712-8693Joan Serra-Sagrista10https://orcid.org/0000-0003-4729-9292David Vilaseca11https://orcid.org/0009-0003-8197-2775Javier Marin12https://orcid.org/0000-0002-6270-2310Satellogic Inc, Barcelona, SpainSatellogic Inc, Barcelona, SpainUniversitat Autònoma de Barcelona - UAB-DEIC-GICI, Bellaterra, SpainEURECAT - Multimedia Technologies Unit, Barcelona, SpainSatellogic Inc, Barcelona, SpainSatellogic Inc, Barcelona, SpainSatellogic Inc, Barcelona, SpainSatellogic Inc, Barcelona, SpainEURECAT - Multimedia Technologies Unit, Barcelona, SpainEURECAT - Multimedia Technologies Unit, Barcelona, SpainUniversitat Autònoma de Barcelona - UAB-DEIC-GICI, Bellaterra, SpainSatellogic Inc, Barcelona, SpainSatellogic Inc, Barcelona, Spain<sc>iquaflow</sc> is a framework that provides a set of tools to assess image quality. The user can add custom metrics that can be easily integrated and a set of unsupervised methods is offered by default. Furthermore, <sc>iquaflow</sc> measures quality by using the performance of AI models trained on the images as a proxy. This also helps to easily make studies of performance degradation of several modifications of the original dataset, for instance, with images reconstructed after different levels of lossy compression; satellite images would be a use case example, since they are commonly compressed before downloading to the ground. In this situation, the optimization problem involves finding images that, while being compressed to their smallest possible file size, still maintain sufficient quality to meet the required performance of the deep learning algorithms. Thus, a study with <sc>iquaflow</sc> is suitable for such case. All this development is wrapped in <sc>Mlflow</sc>: an interactive tool used to visualize and summarize the results. This document describes different use cases and provides links to their respective repositories. To ease the creation of new studies, we include a cookiecutter repository. The source code, issue tracker and aforementioned repositories are all hosted on GitHub.https://ieeexplore.ieee.org/document/10356628/Artificial intelligencedata compressionimage analysisimage processingimage resolutionimage segmentation |
| spellingShingle | Pau Galles Katalin Takats Miguel Hernandez-Cabronero David Berga Luciano Pega Laura Riordan-Chen Clara Garcia Guillermo Becker Adan Garriga Anica Bukva Joan Serra-Sagrista David Vilaseca Javier Marin A New Framework for Evaluating Image Quality Including Deep Learning Task Performances as a Proxy IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing Artificial intelligence data compression image analysis image processing image resolution image segmentation |
| title | A New Framework for Evaluating Image Quality Including Deep Learning Task Performances as a Proxy |
| title_full | A New Framework for Evaluating Image Quality Including Deep Learning Task Performances as a Proxy |
| title_fullStr | A New Framework for Evaluating Image Quality Including Deep Learning Task Performances as a Proxy |
| title_full_unstemmed | A New Framework for Evaluating Image Quality Including Deep Learning Task Performances as a Proxy |
| title_short | A New Framework for Evaluating Image Quality Including Deep Learning Task Performances as a Proxy |
| title_sort | new framework for evaluating image quality including deep learning task performances as a proxy |
| topic | Artificial intelligence data compression image analysis image processing image resolution image segmentation |
| url | https://ieeexplore.ieee.org/document/10356628/ |
| work_keys_str_mv | AT paugalles anewframeworkforevaluatingimagequalityincludingdeeplearningtaskperformancesasaproxy AT katalintakats anewframeworkforevaluatingimagequalityincludingdeeplearningtaskperformancesasaproxy AT miguelhernandezcabronero anewframeworkforevaluatingimagequalityincludingdeeplearningtaskperformancesasaproxy AT davidberga anewframeworkforevaluatingimagequalityincludingdeeplearningtaskperformancesasaproxy AT lucianopega anewframeworkforevaluatingimagequalityincludingdeeplearningtaskperformancesasaproxy AT laurariordanchen anewframeworkforevaluatingimagequalityincludingdeeplearningtaskperformancesasaproxy AT claragarcia anewframeworkforevaluatingimagequalityincludingdeeplearningtaskperformancesasaproxy AT guillermobecker anewframeworkforevaluatingimagequalityincludingdeeplearningtaskperformancesasaproxy AT adangarriga anewframeworkforevaluatingimagequalityincludingdeeplearningtaskperformancesasaproxy AT anicabukva anewframeworkforevaluatingimagequalityincludingdeeplearningtaskperformancesasaproxy AT joanserrasagrista anewframeworkforevaluatingimagequalityincludingdeeplearningtaskperformancesasaproxy AT davidvilaseca anewframeworkforevaluatingimagequalityincludingdeeplearningtaskperformancesasaproxy AT javiermarin anewframeworkforevaluatingimagequalityincludingdeeplearningtaskperformancesasaproxy AT paugalles newframeworkforevaluatingimagequalityincludingdeeplearningtaskperformancesasaproxy AT katalintakats newframeworkforevaluatingimagequalityincludingdeeplearningtaskperformancesasaproxy AT miguelhernandezcabronero newframeworkforevaluatingimagequalityincludingdeeplearningtaskperformancesasaproxy AT davidberga newframeworkforevaluatingimagequalityincludingdeeplearningtaskperformancesasaproxy AT lucianopega newframeworkforevaluatingimagequalityincludingdeeplearningtaskperformancesasaproxy AT laurariordanchen newframeworkforevaluatingimagequalityincludingdeeplearningtaskperformancesasaproxy AT claragarcia newframeworkforevaluatingimagequalityincludingdeeplearningtaskperformancesasaproxy AT guillermobecker newframeworkforevaluatingimagequalityincludingdeeplearningtaskperformancesasaproxy AT adangarriga newframeworkforevaluatingimagequalityincludingdeeplearningtaskperformancesasaproxy AT anicabukva newframeworkforevaluatingimagequalityincludingdeeplearningtaskperformancesasaproxy AT joanserrasagrista newframeworkforevaluatingimagequalityincludingdeeplearningtaskperformancesasaproxy AT davidvilaseca newframeworkforevaluatingimagequalityincludingdeeplearningtaskperformancesasaproxy AT javiermarin newframeworkforevaluatingimagequalityincludingdeeplearningtaskperformancesasaproxy |