Towards Learning Object Detectors with Limited Data for Industrial Applications

In this dissertation, three novel Generalized Few-Shot Object Detection (G-FSOD) approaches are presented to minimize the forgetting of previously learned classes while learning new classes with limited data. The first two approaches reduce the forgetting of base classes if they are still available...

Celý popis

Uloženo v:
Podrobná bibliografie
Hlavní autor: Guirguis, Karim
Médium: Online
Jazyk:angličtina
Vydáno: KIT Scientific Publishing 2025
Témata:
On-line přístup:https://library.oapen.org/handle/20.500.12657/100728
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
_version_ 1869522104050778112
author Guirguis, Karim
author_browse Guirguis, Karim
author_facet Guirguis, Karim
author_sort Guirguis, Karim
collection Directory of Open Access Books
description In this dissertation, three novel Generalized Few-Shot Object Detection (G-FSOD) approaches are presented to minimize the forgetting of previously learned classes while learning new classes with limited data. The first two approaches reduce the forgetting of base classes if they are still available during training. The third approach, for scenarios without base data, uses knowledge distillation to improve the knowledge transfer.
format Online
id doab-20.500.12854ir-158411
institution Directory of Open Access Books
language eng
publishDate 2025
publishDateRange 2025
publishDateSort 2025
publisher KIT Scientific Publishing
publisherStr KIT Scientific Publishing
record_format ojs
spelling doab-20.500.12854ir-1584112025-05-27T05:22:37Z Towards Learning Object Detectors with Limited Data for Industrial Applications Guirguis, Karim Optical Inspection; Object Detection; Deep Learning; Few Shot Learning; Optische Inspektion; Objekt-Erkennung; Computer Vision; Tiefes Lernen thema EDItEUR::U Computing and Information Technology::UY Computer science::UYA Mathematical theory of computation::UYAM Maths for computer scientists In this dissertation, three novel Generalized Few-Shot Object Detection (G-FSOD) approaches are presented to minimize the forgetting of previously learned classes while learning new classes with limited data. The first two approaches reduce the forgetting of base classes if they are still available during training. The third approach, for scenarios without base data, uses knowledge distillation to improve the knowledge transfer. 2025-04-15T04:05:56Z 2025-04-15T04:05:56Z 2025-04-14T12:06:32Z 2025 book https://library.oapen.org/handle/20.500.12657/100728 9783731513896 https://directory.doabooks.org/handle/20.500.12854/158411 eng Schriftenreihe Automatische Sichtprüfung und Bildverarbeitung open access image/jpeg image/jpeg Attribution 4.0 International Attribution 4.0 International https://library.oapen.org/bitstream/20.500.12657/100728/1/towards-learning-object-detectors-with-limited-data-for-industrial-applications.pdf https://library.oapen.org/bitstream/20.500.12657/100728/1/towards-learning-object-detectors-with-limited-data-for-industrial-applications.pdf KIT Scientific Publishing 10.5445/KSP/1000174849 10.5445/KSP/1000174849 68fffc18-8f7b-44fa-ac7e-0b7d7d979bd2 9783731513896 262 open access
spellingShingle Optical Inspection; Object Detection; Deep Learning; Few Shot Learning; Optische Inspektion; Objekt-Erkennung; Computer Vision; Tiefes Lernen
thema EDItEUR::U Computing and Information Technology::UY Computer science::UYA Mathematical theory of computation::UYAM Maths for computer scientists
Guirguis, Karim
Towards Learning Object Detectors with Limited Data for Industrial Applications
title Towards Learning Object Detectors with Limited Data for Industrial Applications
title_full Towards Learning Object Detectors with Limited Data for Industrial Applications
title_fullStr Towards Learning Object Detectors with Limited Data for Industrial Applications
title_full_unstemmed Towards Learning Object Detectors with Limited Data for Industrial Applications
title_short Towards Learning Object Detectors with Limited Data for Industrial Applications
title_sort towards learning object detectors with limited data for industrial applications
topic Optical Inspection; Object Detection; Deep Learning; Few Shot Learning; Optische Inspektion; Objekt-Erkennung; Computer Vision; Tiefes Lernen
thema EDItEUR::U Computing and Information Technology::UY Computer science::UYA Mathematical theory of computation::UYAM Maths for computer scientists
topic_facet Optical Inspection; Object Detection; Deep Learning; Few Shot Learning; Optische Inspektion; Objekt-Erkennung; Computer Vision; Tiefes Lernen
thema EDItEUR::U Computing and Information Technology::UY Computer science::UYA Mathematical theory of computation::UYAM Maths for computer scientists
url https://library.oapen.org/handle/20.500.12657/100728
work_keys_str_mv AT guirguiskarim towardslearningobjectdetectorswithlimiteddataforindustrialapplications