Principles of Catastrophic Forgetting for Continual Semantic Segmentation in Automated Driving

Deep learning excels at extracting complex patterns but faces catastrophic forgetting when fine-tuned on new data. This book investigates how class- and domain-incremental learning affect neural networks for automated driving, identifying semantic shifts and feature changes as key factors. Tools for...

Full beskrivning

Sparad:
Bibliografiska uppgifter
Huvudupphov: Kalb, Tobias Michael
Materialtyp: Online
Språk:engelska
Utgiven: KIT Scientific Publishing 2024
Ämnen:
Länkar:https://library.oapen.org/handle/20.500.12657/94140
Taggar: Lägg till en tagg
Inga taggar, Lägg till första taggen!
_version_ 1869531233245986816
author Kalb, Tobias Michael
author_browse Kalb, Tobias Michael
author_facet Kalb, Tobias Michael
author_sort Kalb, Tobias Michael
collection Directory of Open Access Books
description Deep learning excels at extracting complex patterns but faces catastrophic forgetting when fine-tuned on new data. This book investigates how class- and domain-incremental learning affect neural networks for automated driving, identifying semantic shifts and feature changes as key factors. Tools for quantitatively measuring forgetting are selected and used to show how strategies like image augmentation, pretraining, and architectural adaptations mitigate catastrophic forgetting.
format Online
id doab-20.500.12854ir-147393
institution Directory of Open Access Books
language eng
publishDate 2024
publishDateRange 2024
publishDateSort 2024
publisher KIT Scientific Publishing
publisherStr KIT Scientific Publishing
record_format ojs
spelling doab-20.500.12854ir-1473932025-05-27T05:10:05Z Principles of Catastrophic Forgetting for Continual Semantic Segmentation in Automated Driving Kalb, Tobias Michael Automated Driving; Semantic Segmentation; Catastrophic Forgetting; Continual Learning; Deep Learning; Automatisiertes Fahren; Semantische Segmentierung; Katastrophales Vergessen; Kontinuierliches Lernen thema EDItEUR::U Computing and Information Technology::UY Computer science::UYA Mathematical theory of computation::UYAM Maths for computer scientists Deep learning excels at extracting complex patterns but faces catastrophic forgetting when fine-tuned on new data. This book investigates how class- and domain-incremental learning affect neural networks for automated driving, identifying semantic shifts and feature changes as key factors. Tools for quantitatively measuring forgetting are selected and used to show how strategies like image augmentation, pretraining, and architectural adaptations mitigate catastrophic forgetting. 2024-11-01T04:06:12Z 2024-11-01T04:06:12Z 2024-10-31T14:03:26Z 2024 book https://library.oapen.org/handle/20.500.12657/94140 9783731513735 https://directory.doabooks.org/handle/20.500.12854/147393 eng Karlsruher Schriften zur Anthropomatik open access image/jpeg image/jpeg image/jpeg image/jpeg Attribution-ShareAlike 4.0 International Attribution-ShareAlike 4.0 International Attribution-ShareAlike 4.0 International Attribution-ShareAlike 4.0 International https://library.oapen.org/bitstream/20.500.12657/94140/1/principles-of-catastrophic-forgetting-for-continual-semantic-segmentation-in-automated-driving.pdf https://library.oapen.org/bitstream/20.500.12657/94140/1/principles-of-catastrophic-forgetting-for-continual-semantic-segmentation-in-automated-driving.pdf https://library.oapen.org/bitstream/20.500.12657/94140/1/principles-of-catastrophic-forgetting-for-continual-semantic-segmentation-in-automated-driving.pdf https://library.oapen.org/bitstream/20.500.12657/94140/1/principles-of-catastrophic-forgetting-for-continual-semantic-segmentation-in-automated-driving.pdf KIT Scientific Publishing 10.5445/KSP/1000171902 10.5445/KSP/1000171902 68fffc18-8f7b-44fa-ac7e-0b7d7d979bd2 9783731513735 AG Universitätsverlage 236 open access
spellingShingle Automated Driving; Semantic Segmentation; Catastrophic Forgetting; Continual Learning; Deep Learning; Automatisiertes Fahren; Semantische Segmentierung; Katastrophales Vergessen; Kontinuierliches Lernen
thema EDItEUR::U Computing and Information Technology::UY Computer science::UYA Mathematical theory of computation::UYAM Maths for computer scientists
Kalb, Tobias Michael
Principles of Catastrophic Forgetting for Continual Semantic Segmentation in Automated Driving
title Principles of Catastrophic Forgetting for Continual Semantic Segmentation in Automated Driving
title_full Principles of Catastrophic Forgetting for Continual Semantic Segmentation in Automated Driving
title_fullStr Principles of Catastrophic Forgetting for Continual Semantic Segmentation in Automated Driving
title_full_unstemmed Principles of Catastrophic Forgetting for Continual Semantic Segmentation in Automated Driving
title_short Principles of Catastrophic Forgetting for Continual Semantic Segmentation in Automated Driving
title_sort principles of catastrophic forgetting for continual semantic segmentation in automated driving
topic Automated Driving; Semantic Segmentation; Catastrophic Forgetting; Continual Learning; Deep Learning; Automatisiertes Fahren; Semantische Segmentierung; Katastrophales Vergessen; Kontinuierliches Lernen
thema EDItEUR::U Computing and Information Technology::UY Computer science::UYA Mathematical theory of computation::UYAM Maths for computer scientists
topic_facet Automated Driving; Semantic Segmentation; Catastrophic Forgetting; Continual Learning; Deep Learning; Automatisiertes Fahren; Semantische Segmentierung; Katastrophales Vergessen; Kontinuierliches 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/20.500.12657/94140
work_keys_str_mv AT kalbtobiasmichael principlesofcatastrophicforgettingforcontinualsemanticsegmentationinautomateddriving