Investigating the optimisation of real-world and synthetic object detection training datasets through the consideration of environmental and simulation factors
•Optimisation of a training dataset for training a deep neural network.•Identification of optimal factor levels using ANOVA and Taguchi.•Application of methodology to real-world and synthetic datasets.•Identification of optimal environmental factors for developing simulation environments. Computer v...
Saved in:
| Published in | Intelligent systems with applications Vol. 14; p. 200079 |
|---|---|
| Main Authors | , , , |
| Format | Journal Article |
| Language | English |
| Published |
Elsevier Ltd
01.05.2022
Elsevier |
| Subjects | |
| Online Access | Get full text |
Cover
Loading…
| Abstract | •Optimisation of a training dataset for training a deep neural network.•Identification of optimal factor levels using ANOVA and Taguchi.•Application of methodology to real-world and synthetic datasets.•Identification of optimal environmental factors for developing simulation environments.
Computer vision is used for many industrial applications involving automation, especially those related to efficiency and safety. Computer vision techniques which use machine learning, such as object detectors, need a dataset of images for training and testing. Publicly available datasets or new datasets can be used. However, these datasets rarely consider whether the dataset is leading to optimal performance. Environmental factors, such as lighting and occlusion, will alter the appearance of the images and so images taken under certain condition may have different effects on training. A knowledge gap has formed as to how the test performance of deep neural networks can be improved by considering the effect and interactions of factors where either real or synthetic images are used. The following research illustrates that the different factors can have a significant impact on the test performance and demonstrates a process that can be used on real-world and synthetic images to identify the effect of each factor and discusses how this information may be used to create an optimal training dataset. |
|---|---|
| AbstractList | Computer vision is used for many industrial applications involving automation, especially those related to efficiency and safety. Computer vision techniques which use machine learning, such as object detectors, need a dataset of images for training and testing. Publicly available datasets or new datasets can be used. However, these datasets rarely consider whether the dataset is leading to optimal performance. Environmental factors, such as lighting and occlusion, will alter the appearance of the images and so images taken under certain condition may have different effects on training. A knowledge gap has formed as to how the test performance of deep neural networks can be improved by considering the effect and interactions of factors where either real or synthetic images are used. The following research illustrates that the different factors can have a significant impact on the test performance and demonstrates a process that can be used on real-world and synthetic images to identify the effect of each factor and discusses how this information may be used to create an optimal training dataset. •Optimisation of a training dataset for training a deep neural network.•Identification of optimal factor levels using ANOVA and Taguchi.•Application of methodology to real-world and synthetic datasets.•Identification of optimal environmental factors for developing simulation environments. Computer vision is used for many industrial applications involving automation, especially those related to efficiency and safety. Computer vision techniques which use machine learning, such as object detectors, need a dataset of images for training and testing. Publicly available datasets or new datasets can be used. However, these datasets rarely consider whether the dataset is leading to optimal performance. Environmental factors, such as lighting and occlusion, will alter the appearance of the images and so images taken under certain condition may have different effects on training. A knowledge gap has formed as to how the test performance of deep neural networks can be improved by considering the effect and interactions of factors where either real or synthetic images are used. The following research illustrates that the different factors can have a significant impact on the test performance and demonstrates a process that can be used on real-world and synthetic images to identify the effect of each factor and discusses how this information may be used to create an optimal training dataset. |
| ArticleNumber | 200079 |
| Author | Newman, Callum Justham, Laura Petzing, Jon Goh, Yee Mey |
| Author_xml | – sequence: 1 givenname: Callum orcidid: 0000-0002-7543-6197 surname: Newman fullname: Newman, Callum email: c.newman@lboro.ac.uk – sequence: 2 givenname: Jon orcidid: 0000-0001-5765-5188 surname: Petzing fullname: Petzing, Jon – sequence: 3 givenname: Yee Mey surname: Goh fullname: Goh, Yee Mey – sequence: 4 givenname: Laura orcidid: 0000-0003-1052-1873 surname: Justham fullname: Justham, Laura |
| BookMark | eNp9kd1KJDEQhYMorH8vsFf9Aj3mb5Jp8EbEXQeEvdHrUJ1Uj2l6EkniiE-zr7ppW8SrvUmFqnM-Ujln5DjEgIT8ZHTFKFNX48rnN1hxynk9KNXdETnlSulW0LU4_nb_QS5zHquEbxgTUp6Sv9twwFz8DooPu6Y8YxNfit_7XBsxNHFoEsLUvsU0uQaCa_J7qKribRP7EW1pHJZaZnFJ4MOMcVAgY8mVl-Lr7vmDa2PI3mH6AmM4-BTDHkOBaWH7_eu0zAewJaZ8QU4GmDJeftZz8vTr7vH2vn3483t7e_PQWqFYaRlTSlHeO6HcRvYDVwy51YOkIDvNKWy0ZWvH6thKZJ0cBBMOeiWBiq5X4pxsF66LMJqX5PeQ3k0Ebz4aMe0MpLr0hAbXvFo6hqCU5J3YoIaBStr3moLWtrL4wrIp5pxw-OIxaubEzGjmxMycmFkSq6brxYR1y4PHZLL1GCw6n-r31mf4_9n_AWidpQo |
| Cites_doi | 10.1016/j.autcon.2014.07.006 10.1109/TPAMI.2015.2437384 10.7717/peerj-cs.222 10.1016/j.patrec.2008.04.005 10.1016/j.autcon.2019.04.006 10.1109/TNNLS.2014.2330900 10.1016/j.marpolbul.2018.04.045 10.1007/s11263-014-0733-5 10.1109/TPAMI.2009.167 10.1109/TPAMI.2016.2577031 10.1007/s11263-018-1070-x 10.1016/j.autcon.2015.10.002 10.3389/fbuil.2020.00097 |
| ContentType | Journal Article |
| Copyright | 2022 The Author(s) |
| Copyright_xml | – notice: 2022 The Author(s) |
| DBID | 6I. AAFTH AAYXX CITATION DOA |
| DOI | 10.1016/j.iswa.2022.200079 |
| DatabaseName | ScienceDirect Open Access Titles Elsevier:ScienceDirect:Open Access CrossRef DOAJ Directory of Open Access Journals |
| DatabaseTitle | CrossRef |
| DatabaseTitleList | |
| Database_xml | – sequence: 1 dbid: DOA name: DOAJ Directory of Open Access Journals url: https://www.doaj.org/ sourceTypes: Open Website |
| DeliveryMethod | fulltext_linktorsrc |
| Discipline | Computer Science |
| EISSN | 2667-3053 |
| ExternalDocumentID | oai_doaj_org_article_e5239b91ea6642938e7af040bb70a77c 10_1016_j_iswa_2022_200079 S2667305322000199 |
| GroupedDBID | 6I. AAFTH AAXUO ALMA_UNASSIGNED_HOLDINGS AMRAJ FDB GROUPED_DOAJ M41 M~E ROL 0R~ 0SF AALRI AAYXX ADVLN AFJKZ AITUG CITATION |
| ID | FETCH-LOGICAL-c361t-1166602bd36d84bf261e2c7f40a49720a87c15d136dc4e194f313dab64a039b63 |
| IEDL.DBID | DOA |
| ISSN | 2667-3053 |
| IngestDate | Tue Oct 22 15:13:56 EDT 2024 Thu Sep 26 19:51:11 EDT 2024 Tue Jul 25 20:57:56 EDT 2023 |
| IsDoiOpenAccess | true |
| IsOpenAccess | true |
| IsPeerReviewed | true |
| IsScholarly | true |
| Keywords | Computer vision Autonomous detection Optimisation Object detection Machine learning Synthetic data |
| Language | English |
| License | This is an open access article under the CC BY license. |
| LinkModel | DirectLink |
| MergedId | FETCHMERGED-LOGICAL-c361t-1166602bd36d84bf261e2c7f40a49720a87c15d136dc4e194f313dab64a039b63 |
| ORCID | 0000-0003-1052-1873 0000-0001-5765-5188 0000-0002-7543-6197 |
| OpenAccessLink | https://doaj.org/article/e5239b91ea6642938e7af040bb70a77c |
| ParticipantIDs | doaj_primary_oai_doaj_org_article_e5239b91ea6642938e7af040bb70a77c crossref_primary_10_1016_j_iswa_2022_200079 elsevier_sciencedirect_doi_10_1016_j_iswa_2022_200079 |
| PublicationCentury | 2000 |
| PublicationDate | May 2022 2022-05-00 2022-05-01 |
| PublicationDateYYYYMMDD | 2022-05-01 |
| PublicationDate_xml | – month: 05 year: 2022 text: May 2022 |
| PublicationDecade | 2020 |
| PublicationTitle | Intelligent systems with applications |
| PublicationYear | 2022 |
| Publisher | Elsevier Ltd Elsevier |
| Publisher_xml | – name: Elsevier Ltd – name: Elsevier |
| References | Howal, Jadhav, Arthshi, Nalavade, Shinde (bib0015) 2019 Georgakis, Mousavian, Berg, Kosecka (bib0009) 2017 Deng, Dong, Socher, Li, Li, Fei-Fei (bib0005) 2009 MathsWorks. (n.d.). Johnson-Roberson, Barto, Mehta, Sridhar, Rosaen, Vasudevan (bib0016) 2017 (MathWorks) Retrieved September 9, (2021)., from Girshick, Donahue, Darrell, Malik (bib0011) 2015; 38 Soltani, Zhu, Hammad (bib0036) 2016; 62 Nath, Behzadan (bib0023) 2020 Ros, Sellart, Materzynska, Vazquez, Lopez (bib0033) 2016 Blender. (2019). Newman, Petzing, Goh, Justham (bib0024) 2021; 15 Tsirikoglou, A., Kronander, J., Wrenninge, M., & Unger, J. (2017). Tajeen, Zhu (bib0038) 2014; 48 (Blender) Retrieved April 12, 2021, from . Redmon, Farhadi (bib0028) 2017 Redmon, J., & Farhadi, A. (2018). Redmon, Divvala, Girshick, Farhadi (bib0027) 2016 Ren, He, Girshick, Sun (bib0030) 2017; 39 Unity. (2020). Unity User Manual 2020.1 (Unity Technologies) Retrieved April 850 12, 2021. Wong, Kunii, Ong, Kroupa, Koller (bib0042) 2019 Dwibedi, Misra, Hebert (bib0006) 2017 Tremblay, To, Birchfield (bib0039) 2018 Retrieved April 7, 2021, from (Cortexica) Retrieved April 7, 2021, from Shoa, Zhu, Li (bib0035) 2015; 26 Cordts (bib0004) 2016 Roberts, Golparvar-Fard (bib0032) 2019; 105 Szegedy, Liu, Jia, Sermanet, Reed, Anguelov (bib0037) 2015 (Mathworks) Retrieved April 9, (2021)., from Nowruzi, F.E., .Kapoor, P., Kolhatkar, D., Hassanat, F.A., .Laganiere, R., & Rebut, J. (2019). Felzenswalb, Girshick, McAllester, Ramanan (bib0008) 2010; 32 Martin, Parkes, Zhang, Zhang, McCabe, Duarte (bib0020) 2018; 131 Alhaija, Mustikovela, Mescheder, Geiger, Rother (bib0001) 2018; 126 Peniak, M.. (n.d.). Hinterstoisser, Lepetit, Wohlhart, Konolige (bib0013) 2018 Brostow, Fauqueur, Cipolla (bib0003) 2009; 30 Liu, Anguelov, Erhan, Szegedy, Reed, Fu (bib0019) 2016 Richter, Hayder, Koltun (bib0031) 2017 Wrenninge, M., & Unger, J. (2018). Girshick (bib0010) 2015 Everingham, Ali Eslami, Gool, Williams, Winn, Zisserman (bib0007) 2015; 111 Retrieved April 2021, 09, from Hinterstoisser, Pauly, Heibel, Martina, Bokeloh (bib0014) 2019 He, Zhang, Ren, Sun (bib0012) 2016 Lin, Maire, Belongie, Hays, Perona, Ramanan (bib0018) 2014 Woolf, Fraley, Oom, Terrien, Zalewski (bib0043) 2009 Shafaei, Little, Schmidt (bib0034) 2016 Lin, Goyal, Girshick, He, Dollár (bib0017) 2017 Martin (10.1016/j.iswa.2022.200079_bib0020) 2018; 131 Ros (10.1016/j.iswa.2022.200079_bib0033) 2016 Lin (10.1016/j.iswa.2022.200079_bib0017) 2017 Cordts (10.1016/j.iswa.2022.200079_bib0004) 2016 Everingham (10.1016/j.iswa.2022.200079_bib0007) 2015; 111 Roberts (10.1016/j.iswa.2022.200079_bib0032) 2019; 105 Tajeen (10.1016/j.iswa.2022.200079_bib0038) 2014; 48 Liu (10.1016/j.iswa.2022.200079_bib0019) 2016 Redmon (10.1016/j.iswa.2022.200079_bib0027) 2016 Soltani (10.1016/j.iswa.2022.200079_bib0036) 2016; 62 Deng (10.1016/j.iswa.2022.200079_bib0005) 2009 Redmon (10.1016/j.iswa.2022.200079_bib0028) 2017 Ren (10.1016/j.iswa.2022.200079_bib0030) 2017; 39 Girshick (10.1016/j.iswa.2022.200079_bib0011) 2015; 38 Nath (10.1016/j.iswa.2022.200079_bib0023) 2020 Hinterstoisser (10.1016/j.iswa.2022.200079_bib0013) 2018 Szegedy (10.1016/j.iswa.2022.200079_bib0037) 2015 Tremblay (10.1016/j.iswa.2022.200079_bib0039) 2018 10.1016/j.iswa.2022.200079_bib0029 Lin (10.1016/j.iswa.2022.200079_bib0018) 2014 Johnson-Roberson (10.1016/j.iswa.2022.200079_bib0016) 2017 Hinterstoisser (10.1016/j.iswa.2022.200079_bib0014) 2019 10.1016/j.iswa.2022.200079_bib0022 10.1016/j.iswa.2022.200079_bib0044 Newman (10.1016/j.iswa.2022.200079_bib0024) 2021; 15 Shafaei (10.1016/j.iswa.2022.200079_bib0034) 2016 Woolf (10.1016/j.iswa.2022.200079_bib0043) 2009 Felzenswalb (10.1016/j.iswa.2022.200079_bib0008) 2010; 32 10.1016/j.iswa.2022.200079_bib0021 10.1016/j.iswa.2022.200079_bib0026 Shoa (10.1016/j.iswa.2022.200079_bib0035) 2015; 26 10.1016/j.iswa.2022.200079_bib0002 Richter (10.1016/j.iswa.2022.200079_bib0031) 2017 10.1016/j.iswa.2022.200079_bib0025 Girshick (10.1016/j.iswa.2022.200079_bib0010) 2015 Dwibedi (10.1016/j.iswa.2022.200079_bib0006) 2017 10.1016/j.iswa.2022.200079_bib0040 10.1016/j.iswa.2022.200079_bib0041 Georgakis (10.1016/j.iswa.2022.200079_bib0009) 2017 Brostow (10.1016/j.iswa.2022.200079_bib0003) 2009; 30 He (10.1016/j.iswa.2022.200079_bib0012) 2016 Wong (10.1016/j.iswa.2022.200079_bib0042) 2019 Alhaija (10.1016/j.iswa.2022.200079_bib0001) 2018; 126 Howal (10.1016/j.iswa.2022.200079_bib0015) 2019 |
| References_xml | – volume: 105 year: 2019 ident: bib0032 article-title: End-to-end vision-based detection, tracking and activity analysis of earthmoving equipment filmed at ground level publication-title: Automation in Construction contributor: fullname: Golparvar-Fard – volume: 15 start-page: 15 year: 2021 end-page: 20 ident: bib0024 article-title: Developement of an optimised dataset for training a deep neural network publication-title: Advances in Transdisciplinary Engineering contributor: fullname: Justham – year: 2016 ident: bib0027 article-title: You only look once: Unified, real-time object detection publication-title: Proceedings of the IEEE conference on computer vision and pattern recognition contributor: fullname: Farhadi – year: 2015 ident: bib0010 article-title: Fast R-CNN publication-title: Proceedings of the IEEE international conference on computer vision contributor: fullname: Girshick – volume: 111 start-page: 98 year: 2015 end-page: 136 ident: bib0007 article-title: The PASCAL visual object classes challenge: A retrospective publication-title: International Journal of Computer Vision contributor: fullname: Zisserman – year: 2017 ident: bib0028 article-title: YOLO9000: Better, faster, stronger publication-title: Proceedings of the IEEE conference on computer vision and pattern recognition contributor: fullname: Farhadi – year: 2016 ident: bib0012 article-title: Deep residual learning for image recognition publication-title: Proceedings of the IEEE conference on computer vision and pattern recognition contributor: fullname: Sun – year: 2015 ident: bib0037 article-title: Going deeper with convolutions publication-title: Proceedings of the IEEE conference on computer vision and pattern recognition contributor: fullname: Anguelov – year: 2016 ident: bib0034 article-title: Play and learn: Using video games to train computer vision models publication-title: Proceedings of the British machine vision conference contributor: fullname: Schmidt – year: 2018 ident: bib0013 article-title: On pre-trained features and synthetic images for deep learning publication-title: Proceedings of the European conference on computer vision contributor: fullname: Konolige – year: 2016 ident: bib0019 article-title: SSD: Single shot multibox detector publication-title: Proceedings of the European conference on computer vision contributor: fullname: Fu – start-page: 3213 year: 2016 end-page: 3223 ident: bib0004 article-title: The Cityscapes Dataset for Semantic Urban Scene Understanding publication-title: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) contributor: fullname: Cordts – year: 2020 ident: bib0023 article-title: Deep convolutional networks for construction object detection under different Visual conditions publication-title: Frontiers in Built Environment contributor: fullname: Behzadan – volume: 39 start-page: 1137 year: 2017 end-page: 1149 ident: bib0030 article-title: Faster R-CNN: Towards real-time object detection with region proposal networks publication-title: IEEE Transactions on Pattern Analysis and Machine Intelligence contributor: fullname: Sun – year: 2018 ident: bib0039 article-title: Falling things: A synthetic dataset for 3D object detection and pose estimation publication-title: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition workshops contributor: fullname: Birchfield – year: 2017 ident: bib0016 article-title: Driving in the matrix: Can virtual worlds replace human-generated annotations for real world tasks? publication-title: Proceedings of the IEEE international conference on robotics and automation contributor: fullname: Vasudevan – year: 2009 ident: bib0005 article-title: ImageNet: A large-scale hierarchical image database publication-title: Proceedings of the IEEE conference on computer vision and pattern recognition contributor: fullname: Fei-Fei – volume: 30 start-page: 88 year: 2009 end-page: 97 ident: bib0003 article-title: Semantic object classes in video: A high-definition ground truth database publication-title: Pattern Recognition Letters contributor: fullname: Cipolla – volume: 62 start-page: 14 year: 2016 end-page: 23 ident: bib0036 article-title: Automated annotation for visual recognition of construction resources using synthetic images publication-title: Automation in Construction contributor: fullname: Hammad – year: 2017 ident: bib0017 article-title: Focal loss for dense object detection publication-title: Proceedings of the IEEE international conference on computer vision contributor: fullname: Dollár – year: 2014 ident: bib0018 article-title: Microsoft COCO: Common objects in context publication-title: Proceedings of the European conference on computer vision contributor: fullname: Ramanan – volume: 38 start-page: 142 year: 2015 end-page: 158 ident: bib0011 article-title: Region-based convolutional networks for accurate object detection and segmentation publication-title: IEEE Transactions on Pattern Analysis and Machine Intelligence contributor: fullname: Malik – start-page: 86 year: 2019 end-page: 93 ident: bib0015 article-title: Object detection for autonomous vehicle using tensorflow publication-title: Proceedings of the international conference on intelligent computing, information and control systems contributor: fullname: Shinde – volume: 126 start-page: 961 year: 2018 end-page: 972 ident: bib0001 article-title: Augmented reality meets computer vision: Efficient data generation for urban driving scenes publication-title: International Journal of Computer Vision contributor: fullname: Rother – year: 2016 ident: bib0033 article-title: The SYNTHIA dataset: A large collection of synthetic images for semantic segmentation of urban scenes publication-title: Proceedings of the IEEE conference on computer vision and pattern recognition contributor: fullname: Lopez – volume: 48 start-page: 1 year: 2014 end-page: 10 ident: bib0038 article-title: Image dataset development for measuring construction equipment recognition performance publication-title: Autoamtion in Construction contributor: fullname: Zhu – year: 2019 ident: bib0014 article-title: An annotation saved is an annotation earned: Using fully synthetic training for object detection publication-title: Proceedings of the IEEE/CVF international conference on computer vision workshop contributor: fullname: Bokeloh – volume: 131 start-page: 662 year: 2018 end-page: 673 ident: bib0020 article-title: Use of unmanned aerial vehicles for efficient beach litter monitoring publication-title: Marine Pollution Bulletin contributor: fullname: Duarte – volume: 26 start-page: 1019 year: 2015 end-page: 1034 ident: bib0035 article-title: Transfer learning for visual categorization: A survey publication-title: IEEE Transactions on Neural Networks and Learning Systems contributor: fullname: Li – year: 2017 ident: bib0006 article-title: Cut, paste and learn: Surprisingly easy synthesis for instance detection publication-title: Proceedings of the IEEE international conference on computer vision contributor: fullname: Hebert – year: 2017 ident: bib0009 article-title: Synthesizing training data for object detection in indoor scenes publication-title: Robotics: Science and systems contributor: fullname: Kosecka – year: 2009 ident: bib0043 article-title: Design of experiments via taguchi methods - orthogonal arrays publication-title: Chemical process dynamics and controls contributor: fullname: Zalewski – year: 2017 ident: bib0031 article-title: Playing for Benchmarks publication-title: Proceedings of the international conference on computer vision contributor: fullname: Koltun – volume: 32 start-page: 1627 year: 2010 end-page: 1645 ident: bib0008 article-title: Object detection with discriminatively trained part-based models publication-title: IEEE Transactions on Pattern Analysis and Machine Intelligence contributor: fullname: Ramanan – year: 2019 ident: bib0042 article-title: Synthetic dataset generation for object-to-model deep learning in industrial applications publication-title: PeerJ Computer Science contributor: fullname: Koller – year: 2017 ident: 10.1016/j.iswa.2022.200079_bib0028 article-title: YOLO9000: Better, faster, stronger contributor: fullname: Redmon – year: 2016 ident: 10.1016/j.iswa.2022.200079_bib0012 article-title: Deep residual learning for image recognition contributor: fullname: He – year: 2016 ident: 10.1016/j.iswa.2022.200079_bib0027 article-title: You only look once: Unified, real-time object detection contributor: fullname: Redmon – year: 2018 ident: 10.1016/j.iswa.2022.200079_bib0039 article-title: Falling things: A synthetic dataset for 3D object detection and pose estimation contributor: fullname: Tremblay – year: 2015 ident: 10.1016/j.iswa.2022.200079_bib0037 article-title: Going deeper with convolutions contributor: fullname: Szegedy – volume: 48 start-page: 1 year: 2014 ident: 10.1016/j.iswa.2022.200079_bib0038 article-title: Image dataset development for measuring construction equipment recognition performance publication-title: Autoamtion in Construction doi: 10.1016/j.autcon.2014.07.006 contributor: fullname: Tajeen – volume: 38 start-page: 142 issue: 1 year: 2015 ident: 10.1016/j.iswa.2022.200079_bib0011 article-title: Region-based convolutional networks for accurate object detection and segmentation publication-title: IEEE Transactions on Pattern Analysis and Machine Intelligence doi: 10.1109/TPAMI.2015.2437384 contributor: fullname: Girshick – year: 2017 ident: 10.1016/j.iswa.2022.200079_bib0016 article-title: Driving in the matrix: Can virtual worlds replace human-generated annotations for real world tasks? contributor: fullname: Johnson-Roberson – start-page: 3213 year: 2016 ident: 10.1016/j.iswa.2022.200079_bib0004 article-title: The Cityscapes Dataset for Semantic Urban Scene Understanding contributor: fullname: Cordts – year: 2019 ident: 10.1016/j.iswa.2022.200079_bib0042 article-title: Synthetic dataset generation for object-to-model deep learning in industrial applications publication-title: PeerJ Computer Science doi: 10.7717/peerj-cs.222 contributor: fullname: Wong – year: 2014 ident: 10.1016/j.iswa.2022.200079_bib0018 article-title: Microsoft COCO: Common objects in context contributor: fullname: Lin – ident: 10.1016/j.iswa.2022.200079_bib0044 – ident: 10.1016/j.iswa.2022.200079_bib0025 – ident: 10.1016/j.iswa.2022.200079_bib0002 – ident: 10.1016/j.iswa.2022.200079_bib0021 – volume: 30 start-page: 88 issue: 2 year: 2009 ident: 10.1016/j.iswa.2022.200079_bib0003 article-title: Semantic object classes in video: A high-definition ground truth database publication-title: Pattern Recognition Letters doi: 10.1016/j.patrec.2008.04.005 contributor: fullname: Brostow – volume: 105 year: 2019 ident: 10.1016/j.iswa.2022.200079_bib0032 article-title: End-to-end vision-based detection, tracking and activity analysis of earthmoving equipment filmed at ground level publication-title: Automation in Construction doi: 10.1016/j.autcon.2019.04.006 contributor: fullname: Roberts – ident: 10.1016/j.iswa.2022.200079_bib0040 – ident: 10.1016/j.iswa.2022.200079_bib0029 – year: 2017 ident: 10.1016/j.iswa.2022.200079_bib0031 article-title: Playing for Benchmarks contributor: fullname: Richter – volume: 15 start-page: 15 year: 2021 ident: 10.1016/j.iswa.2022.200079_bib0024 article-title: Developement of an optimised dataset for training a deep neural network publication-title: Advances in Transdisciplinary Engineering contributor: fullname: Newman – volume: 26 start-page: 1019 issue: 5 year: 2015 ident: 10.1016/j.iswa.2022.200079_bib0035 article-title: Transfer learning for visual categorization: A survey publication-title: IEEE Transactions on Neural Networks and Learning Systems doi: 10.1109/TNNLS.2014.2330900 contributor: fullname: Shoa – year: 2019 ident: 10.1016/j.iswa.2022.200079_bib0014 article-title: An annotation saved is an annotation earned: Using fully synthetic training for object detection contributor: fullname: Hinterstoisser – year: 2017 ident: 10.1016/j.iswa.2022.200079_bib0017 article-title: Focal loss for dense object detection contributor: fullname: Lin – year: 2018 ident: 10.1016/j.iswa.2022.200079_bib0013 article-title: On pre-trained features and synthetic images for deep learning contributor: fullname: Hinterstoisser – volume: 131 start-page: 662 issue: Part A year: 2018 ident: 10.1016/j.iswa.2022.200079_bib0020 article-title: Use of unmanned aerial vehicles for efficient beach litter monitoring publication-title: Marine Pollution Bulletin doi: 10.1016/j.marpolbul.2018.04.045 contributor: fullname: Martin – volume: 111 start-page: 98 year: 2015 ident: 10.1016/j.iswa.2022.200079_bib0007 article-title: The PASCAL visual object classes challenge: A retrospective publication-title: International Journal of Computer Vision doi: 10.1007/s11263-014-0733-5 contributor: fullname: Everingham – volume: 32 start-page: 1627 issue: 9 year: 2010 ident: 10.1016/j.iswa.2022.200079_bib0008 article-title: Object detection with discriminatively trained part-based models publication-title: IEEE Transactions on Pattern Analysis and Machine Intelligence doi: 10.1109/TPAMI.2009.167 contributor: fullname: Felzenswalb – year: 2009 ident: 10.1016/j.iswa.2022.200079_bib0005 article-title: ImageNet: A large-scale hierarchical image database contributor: fullname: Deng – year: 2017 ident: 10.1016/j.iswa.2022.200079_bib0009 article-title: Synthesizing training data for object detection in indoor scenes contributor: fullname: Georgakis – ident: 10.1016/j.iswa.2022.200079_bib0022 – volume: 39 start-page: 1137 issue: 6 year: 2017 ident: 10.1016/j.iswa.2022.200079_bib0030 article-title: Faster R-CNN: Towards real-time object detection with region proposal networks publication-title: IEEE Transactions on Pattern Analysis and Machine Intelligence doi: 10.1109/TPAMI.2016.2577031 contributor: fullname: Ren – year: 2016 ident: 10.1016/j.iswa.2022.200079_bib0033 article-title: The SYNTHIA dataset: A large collection of synthetic images for semantic segmentation of urban scenes contributor: fullname: Ros – volume: 126 start-page: 961 year: 2018 ident: 10.1016/j.iswa.2022.200079_bib0001 article-title: Augmented reality meets computer vision: Efficient data generation for urban driving scenes publication-title: International Journal of Computer Vision doi: 10.1007/s11263-018-1070-x contributor: fullname: Alhaija – volume: 62 start-page: 14 year: 2016 ident: 10.1016/j.iswa.2022.200079_bib0036 article-title: Automated annotation for visual recognition of construction resources using synthetic images publication-title: Automation in Construction doi: 10.1016/j.autcon.2015.10.002 contributor: fullname: Soltani – year: 2015 ident: 10.1016/j.iswa.2022.200079_bib0010 article-title: Fast R-CNN contributor: fullname: Girshick – ident: 10.1016/j.iswa.2022.200079_bib0041 – year: 2017 ident: 10.1016/j.iswa.2022.200079_bib0006 article-title: Cut, paste and learn: Surprisingly easy synthesis for instance detection contributor: fullname: Dwibedi – start-page: 86 year: 2019 ident: 10.1016/j.iswa.2022.200079_bib0015 article-title: Object detection for autonomous vehicle using tensorflow contributor: fullname: Howal – ident: 10.1016/j.iswa.2022.200079_bib0026 – year: 2020 ident: 10.1016/j.iswa.2022.200079_bib0023 article-title: Deep convolutional networks for construction object detection under different Visual conditions publication-title: Frontiers in Built Environment doi: 10.3389/fbuil.2020.00097 contributor: fullname: Nath – year: 2009 ident: 10.1016/j.iswa.2022.200079_bib0043 article-title: Design of experiments via taguchi methods - orthogonal arrays contributor: fullname: Woolf – year: 2016 ident: 10.1016/j.iswa.2022.200079_bib0019 article-title: SSD: Single shot multibox detector contributor: fullname: Liu – year: 2016 ident: 10.1016/j.iswa.2022.200079_bib0034 article-title: Play and learn: Using video games to train computer vision models contributor: fullname: Shafaei |
| SSID | ssj0002811344 |
| Score | 2.2426722 |
| Snippet | •Optimisation of a training dataset for training a deep neural network.•Identification of optimal factor levels using ANOVA and Taguchi.•Application of... Computer vision is used for many industrial applications involving automation, especially those related to efficiency and safety. Computer vision techniques... |
| SourceID | doaj crossref elsevier |
| SourceType | Open Website Aggregation Database Publisher |
| StartPage | 200079 |
| SubjectTerms | Autonomous detection Computer vision Machine learning Object detection Optimisation Synthetic data |
| Title | Investigating the optimisation of real-world and synthetic object detection training datasets through the consideration of environmental and simulation factors |
| URI | https://dx.doi.org/10.1016/j.iswa.2022.200079 https://doaj.org/article/e5239b91ea6642938e7af040bb70a77c |
| Volume | 14 |
| hasFullText | 1 |
| inHoldings | 1 |
| isFullTextHit | |
| isPrint | |
| link | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwrV09T8MwELVQJxa-EeVLHthQhJ04tjMCalUhlYlK3SLbcVArmiIShFj4K_xVzo5TZYKFJUNinaPc2ffsnN9D6EozFUvN3R_G1F90pErOI2kNgNtYypK4g8LTRz6ZsYd5Ou9JfbmasJYeuP1wNxZWSpnOqFUcoHKWSCtUCZGntSBKCONnX5L1FlNLv2VEaeKVXCEBwSiCUAsnZtrirkX94UiHYk9FSVwdVy8refL-XnLqJZzxHtoJSBHftm-4j7ZsdYB2OxUGHAblIfruUWVUzxgAHV7DPLAKdTp4XWIAhi-R50bFqipw_VlBKzCL19ptw-DCNr4iq8KdYgR2laO1bWochHy8XRPUPTeGe8fk4FW97cUqCILhIOVzhGbj0dP9JAqyC5FJOG0i6v4kklgXCS8k0yWssWxsRMmIYpmIiZLC0LSg8NgwSzNWJjQplOZMEXAVT47RoFpX9gRhao0UVqZWQbMEJkdRUpsx40lpALkM0XX32fPXll0j78rOlrlzUu6clLdOGqI755lNS8eM7W9AvOQhXvK_4mWI0s6veQAZLXgAU4tfOj_9j87P0LYz2RZMnqNB8_ZuLwDUNPrSxy9cp1-jHybE9rU |
| link.rule.ids | 315,783,787,867,2109,27938,27939 |
| linkProvider | Directory of Open Access Journals |
| openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Investigating+the+optimisation+of+real-world+and+synthetic+object+detection+training+datasets+through+the+consideration+of+environmental+and+simulation+factors&rft.jtitle=Intelligent+systems+with+applications&rft.au=Callum+Newman&rft.au=Jon+Petzing&rft.au=Yee+Mey+Goh&rft.au=Laura+Justham&rft.date=2022-05-01&rft.pub=Elsevier&rft.issn=2667-3053&rft.eissn=2667-3053&rft.volume=14&rft.spage=200079&rft_id=info:doi/10.1016%2Fj.iswa.2022.200079&rft.externalDBID=DOA&rft.externalDocID=oai_doaj_org_article_e5239b91ea6642938e7af040bb70a77c |
| thumbnail_l | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=2667-3053&client=summon |
| thumbnail_m | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=2667-3053&client=summon |
| thumbnail_s | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=2667-3053&client=summon |