Synthetic Data Generation to Speed-Up the Object Recognition Pipeline
This paper provides a methodology for the production of synthetic images for training neural networks to recognise shapes and objects. There are many scenarios in which it is difficult, expensive and even dangerous to produce a set of images that is satisfactory for the training of a neural network....
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-12-01
|
| Series: | Electronics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2079-9292/11/1/2 |
| _version_ | 1797499325211738112 |
|---|---|
| author | Damiano Perri Marco Simonetti Osvaldo Gervasi |
| author_facet | Damiano Perri Marco Simonetti Osvaldo Gervasi |
| author_sort | Damiano Perri |
| collection | DOAJ |
| description | This paper provides a methodology for the production of synthetic images for training neural networks to recognise shapes and objects. There are many scenarios in which it is difficult, expensive and even dangerous to produce a set of images that is satisfactory for the training of a neural network. The development of 3D modelling software has nowadays reached such a level of realism and ease of use that it seemed natural to explore this innovative path and to give an answer regarding the reliability of this method that bases the training of the neural network on synthetic images. The results obtained in the two proposed use cases, that of the recognition of a pictorial style and that of the recognition of men at sea, lead us to support the validity of the approach, provided that the work is conducted in a very scrupulous and rigorous manner, exploiting the full potential of the modelling software. The code produced, which automatically generates the transformations necessary for the data augmentation of each image, and the generation of random environmental conditions in the case of Blender and Unity3D software, is available under the GPL licence on GitHub. The results obtained lead us to affirm that through the good practices presented in the article, we have defined a simple, reliable, economic and safe method to feed the training phase of a neural network dedicated to the recognition of objects and features to be applied to various contexts. |
| first_indexed | 2024-03-10T03:45:47Z |
| format | Article |
| id | doaj.art-cc3e53a15cc84fc98971e9b2ec323a5e |
| institution | Directory Open Access Journal |
| issn | 2079-9292 |
| language | English |
| last_indexed | 2024-03-10T03:45:47Z |
| publishDate | 2021-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Electronics |
| spelling | doaj.art-cc3e53a15cc84fc98971e9b2ec323a5e2023-11-23T11:21:24ZengMDPI AGElectronics2079-92922021-12-01111210.3390/electronics11010002Synthetic Data Generation to Speed-Up the Object Recognition PipelineDamiano Perri0Marco Simonetti1Osvaldo Gervasi2Department of Mathematics and Computer Science, University of Florence, 50134 Florence, ItalyDepartment of Mathematics and Computer Science, University of Florence, 50134 Florence, ItalyDepartment of Mathematics and Computer Science, University of Perugia, 06123 Perugia, ItalyThis paper provides a methodology for the production of synthetic images for training neural networks to recognise shapes and objects. There are many scenarios in which it is difficult, expensive and even dangerous to produce a set of images that is satisfactory for the training of a neural network. The development of 3D modelling software has nowadays reached such a level of realism and ease of use that it seemed natural to explore this innovative path and to give an answer regarding the reliability of this method that bases the training of the neural network on synthetic images. The results obtained in the two proposed use cases, that of the recognition of a pictorial style and that of the recognition of men at sea, lead us to support the validity of the approach, provided that the work is conducted in a very scrupulous and rigorous manner, exploiting the full potential of the modelling software. The code produced, which automatically generates the transformations necessary for the data augmentation of each image, and the generation of random environmental conditions in the case of Blender and Unity3D software, is available under the GPL licence on GitHub. The results obtained lead us to affirm that through the good practices presented in the article, we have defined a simple, reliable, economic and safe method to feed the training phase of a neural network dedicated to the recognition of objects and features to be applied to various contexts.https://www.mdpi.com/2079-9292/11/1/2Unity3DBlendervirtual realitysynthetic dataset generationmachine learningneural networks |
| spellingShingle | Damiano Perri Marco Simonetti Osvaldo Gervasi Synthetic Data Generation to Speed-Up the Object Recognition Pipeline Electronics Unity3D Blender virtual reality synthetic dataset generation machine learning neural networks |
| title | Synthetic Data Generation to Speed-Up the Object Recognition Pipeline |
| title_full | Synthetic Data Generation to Speed-Up the Object Recognition Pipeline |
| title_fullStr | Synthetic Data Generation to Speed-Up the Object Recognition Pipeline |
| title_full_unstemmed | Synthetic Data Generation to Speed-Up the Object Recognition Pipeline |
| title_short | Synthetic Data Generation to Speed-Up the Object Recognition Pipeline |
| title_sort | synthetic data generation to speed up the object recognition pipeline |
| topic | Unity3D Blender virtual reality synthetic dataset generation machine learning neural networks |
| url | https://www.mdpi.com/2079-9292/11/1/2 |
| work_keys_str_mv | AT damianoperri syntheticdatagenerationtospeeduptheobjectrecognitionpipeline AT marcosimonetti syntheticdatagenerationtospeeduptheobjectrecognitionpipeline AT osvaldogervasi syntheticdatagenerationtospeeduptheobjectrecognitionpipeline |