| dc.contributor.advisor | Μαγκλογιάννης, Ηλίας | |
| dc.contributor.advisor | Maglogiannis, Ilias | |
| dc.contributor.author | Πάγκος, Κωνσταντίνος | |
| dc.contributor.author | Pagkos, Konstantinos | |
| dc.date.accessioned | 2023-08-31T08:10:17Z | |
| dc.date.available | 2023-08-31T08:10:17Z | |
| dc.date.issued | 2023-06-15 | |
| dc.identifier.uri | https://dione.lib.unipi.gr/xmlui/handle/unipi/15674 | |
| dc.identifier.uri | http://dx.doi.org/10.26267/unipi_dione/3096 | |
| dc.format.extent | 73 | el |
| dc.language.iso | en | el |
| dc.publisher | Πανεπιστήμιο Πειραιώς | el |
| dc.title | Data processing from sensors at the edge | el |
| dc.type | Master Thesis | el |
| dc.contributor.department | Σχολή Τεχνολογιών Πληροφορικής και Επικοινωνιών. Τμήμα Ψηφιακών Συστημάτων | el |
| dc.description.abstractEN | The purpose of this thesis is to test and examine the capabilities of different state-of-the-art convolutional neural network architectures for edge applications and evaluate their maturity for use in real-time medical applications. For this reason, we utilize a publicly available annotated dataset containing images of the human gastrointestinal tract and the use of one of the most advanced AI edge accelerators. Ultimately, we test, evaluate and compare the performance of several models and provide insight both into the nature of the dataset, as well as into the capacity and potential of the latest advancements on the field of lightweight convolutional neural networks optimized for embedded devices. | el |
| dc.contributor.master | Πληροφοριακά Συστήματα και Υπηρεσίες | el |
| dc.subject.keyword | Convolutional neural network | el |
| dc.subject.keyword | Edge applications | el |
| dc.subject.keyword | Medical image processing | el |
| dc.date.defense | 2023-07-14 | |
