{"id":{"repo_id":"trento","oai_identifier":"oai:iris.unitn.it:11572/481852"},"canonical_url":"https://search.dev.ndltd.org/etd/trento/oai:iris.unitn.it:11572/481852","repository":{"repo_id":"trento","name":"Università degli Studi di Trento","base_url":"https://iris.unitn.it/oai/request"},"display":{"title":"Generalizing Under Data Scarcity. Enhancing the representation capability from few samples.","abstract":"The widespread adoption of deep learning in both research and industrial contexts has revealed a central limitation: many real-world applications lack large, diverse, and reliably labeled datasets. This challenge is particularly evident in domains where data acquisition is costly, error-prone, or inherently scarce, such as industrial inspection, anomaly detection and localization. This thesis investigates how learning systems can be designed to operate effectively when only a small number of samples are available at training or inference time. The first part of the thesis focuses on meta-learning transformers for supervised and unsupervised few-shot tasks. We explore how transformers behave when trained on structured, multi-domain datasets under controlled conditions, where train/test contamination can be explicitly avoided. By reframing few-shot learning as a sequence modeling problem, we analyze the generalization capabilities of in-context learners across domains, and study how training order influences performance. We propose the GEOM framework for supervised few-shot classification and extend its principles to the unsupervised setting with CAMeLU, demonstrating state-of-the-art performance in cross-domain scenarios. The second part of the thesis addresses the gap between academic research and real-world industrial constraints. Working with an Italian company specializing in glass inspection systems, we propose two domain-specific solutions. The first is a few-shot approach for structural glass defect classification, enabling flexible adaptation to new defect types and variations in glass materials. The second is a reconstruction-based anomaly detection pipeline for identifying irregularities in silk-screen printed patterns, where labeled data are extremely scarce. This dissertation highlights the importance of designing models that do not rely on large-scale datasets, but instead leverage task structure, adaptation mechanisms, and data-efficient learning strategies. By bridging foundational research on meta-learning with concrete industrial use cases, the thesis demonstrates that few-shot paradigms can be robust, scalable, and practically applicable in demanding environments.","abstract_html":"The widespread adoption of deep learning in both research and industrial contexts has revealed a central limitation: many real-world applications lack large, diverse, and reliably labeled datasets. This challenge is particularly evident in domains where data acquisition is costly, error-prone, or inherently scarce, such as industrial inspection, anomaly detection and localization. This thesis investigates how learning systems can be designed to operate effectively when only a small number of samples are available at training or inference time. The first part of the thesis focuses on meta-learning transformers for supervised and unsupervised few-shot tasks. We explore how transformers behave when trained on structured, multi-domain datasets under controlled conditions, where train/test contamination can be explicitly avoided. By reframing few-shot learning as a sequence modeling problem, we analyze the generalization capabilities of in-context learners across domains, and study how training order influences performance. We propose the GEOM framework for supervised few-shot classification and extend its principles to the unsupervised setting with CAMeLU, demonstrating state-of-the-art performance in cross-domain scenarios. The second part of the thesis addresses the gap between academic research and real-world industrial constraints. Working with an Italian company specializing in glass inspection systems, we propose two domain-specific solutions. The first is a few-shot approach for structural glass defect classification, enabling flexible adaptation to new defect types and variations in glass materials. The second is a reconstruction-based anomaly detection pipeline for identifying irregularities in silk-screen printed patterns, where labeled data are extremely scarce. This dissertation highlights the importance of designing models that do not rely on large-scale datasets, but instead leverage task structure, adaptation mechanisms, and data-efficient learning strategies. By bridging foundational research on meta-learning with concrete industrial use cases, the thesis demonstrates that few-shot paradigms can be robust, scalable, and practically applicable in demanding environments.","abstract_has_math":false,"creators":["Braccaioli, Lorenzo"],"institution":"Università degli studi di Trento","degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":null,"school":null,"contributors":["co-supervisor: Chiara Corridori","Conci, Nicola"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2026,"date_issued":"2026-04-15","date_published":"2026-04-15","updated_at":"2026-07-24T05:04:12Z","subjects":["few-shot learning","meta-learning","in-context learning","industrial anomaly detection"],"languages":["eng"],"rights":["info:eu-repo/semantics/embargoedAccess","license:Tutti i diritti riservati (All rights reserved)","license uri:iris.PRI01"],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"https://hdl.handle.net/11572/481852","outbound_label":"Handle","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["co-supervisor: Chiara Corridori","Braccaioli, Lorenzo","Conci, Nicola"]},{"key":"dc:creator","label":"Author","values":["Braccaioli, Lorenzo"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2026-04-15"]},{"key":"dc:publisher","label":"Institution","values":["Università degli studi di Trento","place:TRENTO"]},{"key":"dc:relation","label":"Dc Relation","values":["firstpage:1","lastpage:140","numberofpages:140","alleditors:co-supervisor: Chiara Corridori"]},{"key":"dc:type","label":"Dc Type","values":["info:eu-repo/semantics/doctoralThesis"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["few-shot learning","meta-learning","in-context learning","industrial anomaly detection"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["eng"]},{"key":"dc:rights","label":"Dc Rights","values":["info:eu-repo/semantics/embargoedAccess","license:Tutti i diritti riservati (All rights reserved)","license uri:iris.PRI01"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["https://hdl.handle.net/11572/481852"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["The widespread adoption of deep learning in both research and industrial contexts has revealed a central limitation: many real-world applications lack large, diverse, and reliably labeled datasets. This challenge is particularly evident in domains where data acquisition is costly, error-prone, or inherently scarce, such as industrial inspection, anomaly detection and localization. This thesis investigates how learning systems can be designed to operate effectively when only a small number of samples are available at training or inference time. The first part of the thesis focuses on meta-learning transformers for supervised and unsupervised few-shot tasks. We explore how transformers behave when trained on structured, multi-domain datasets under controlled conditions, where train/test contamination can be explicitly avoided. By reframing few-shot learning as a sequence modeling problem, we analyze the generalization capabilities of in-context learners across domains, and study how training order influences performance. We propose the GEOM framework for supervised few-shot classification and extend its principles to the unsupervised setting with CAMeLU, demonstrating state-of-the-art performance in cross-domain scenarios. The second part of the thesis addresses the gap between academic research and real-world industrial constraints. Working with an Italian company specializing in glass inspection systems, we propose two domain-specific solutions. The first is a few-shot approach for structural glass defect classification, enabling flexible adaptation to new defect types and variations in glass materials. The second is a reconstruction-based anomaly detection pipeline for identifying irregularities in silk-screen printed patterns, where labeled data are extremely scarce. This dissertation highlights the importance of designing models that do not rely on large-scale datasets, but instead leverage task structure, adaptation mechanisms, and data-efficient learning strategies. By bridging foundational research on meta-learning with concrete industrial use cases, the thesis demonstrates that few-shot paradigms can be robust, scalable, and practically applicable in demanding environments."]},{"key":"dc:title","label":"Title","values":["Generalizing Under Data Scarcity. Enhancing the representation capability from few samples."]}]}],"canonical_facts":{"dc:contributor":["co-supervisor: Chiara Corridori","Braccaioli, Lorenzo","Conci, Nicola"],"dc:creator":["Braccaioli, Lorenzo"],"dc:date":["2026-04-15"],"dc:description":["The widespread adoption of deep learning in both research and industrial contexts has revealed a central limitation: many real-world applications lack large, diverse, and reliably labeled datasets. This challenge is particularly evident in domains where data acquisition is costly, error-prone, or inherently scarce, such as industrial inspection, anomaly detection and localization. This thesis investigates how learning systems can be designed to operate effectively when only a small number of samples are available at training or inference time. The first part of the thesis focuses on meta-learning transformers for supervised and unsupervised few-shot tasks. We explore how transformers behave when trained on structured, multi-domain datasets under controlled conditions, where train/test contamination can be explicitly avoided. By reframing few-shot learning as a sequence modeling problem, we analyze the generalization capabilities of in-context learners across domains, and study how training order influences performance. We propose the GEOM framework for supervised few-shot classification and extend its principles to the unsupervised setting with CAMeLU, demonstrating state-of-the-art performance in cross-domain scenarios. The second part of the thesis addresses the gap between academic research and real-world industrial constraints. Working with an Italian company specializing in glass inspection systems, we propose two domain-specific solutions. The first is a few-shot approach for structural glass defect classification, enabling flexible adaptation to new defect types and variations in glass materials. The second is a reconstruction-based anomaly detection pipeline for identifying irregularities in silk-screen printed patterns, where labeled data are extremely scarce. This dissertation highlights the importance of designing models that do not rely on large-scale datasets, but instead leverage task structure, adaptation mechanisms, and data-efficient learning strategies. By bridging foundational research on meta-learning with concrete industrial use cases, the thesis demonstrates that few-shot paradigms can be robust, scalable, and practically applicable in demanding environments."],"dc:identifier":["https://hdl.handle.net/11572/481852"],"dc:language":["eng"],"dc:publisher":["Università degli studi di Trento","place:TRENTO"],"dc:relation":["firstpage:1","lastpage:140","numberofpages:140","alleditors:co-supervisor: Chiara Corridori"],"dc:rights":["info:eu-repo/semantics/embargoedAccess","license:Tutti i diritti riservati (All rights reserved)","license uri:iris.PRI01"],"dc:subject":["few-shot learning","meta-learning","in-context learning","industrial anomaly detection"],"dc:title":["Generalizing Under Data Scarcity. Enhancing the representation capability from few samples."],"dc:type":["info:eu-repo/semantics/doctoralThesis"]},"updated_at":"2026-07-24T05:04:12Z"}