Open-Set Text Recognition : Concepts, Framework, and Algorithms

• Autore: Yin Xu-Cheng
• Pubblicazione: Singapore : , : Springer Singapore Pte. Limited, , 2024; ©2024
• Edizione: 1st ed.
• Descrizione fisica: 1 online resource (130 pages)
• Altri autori: YangChun LiuChang
• Nota di contenuto: Intro -- Preface -- Contents -- Acronyms -- 1 Introduction -- 1.1 Introduction -- References -- 2 Background -- 2.1 Agents in Open World -- 2.1.1 Open-World Learning Tasks -- 2.1.2 Open-World Learning Approaches -- 2.1.3 Typical Applications -- 2.2 Text Recognition -- 2.2.1 Conventional Close-Set Text Recognition -- 2.2.2 Beyond Close-Set Text Recognition -- References -- 3 Open-Set Text Recognition: Concept, Dataset, Protocol, and Framework -- 3.1 Concept -- 3.1.1 Aims, Goals, and Scope -- 3.1.2 Task Definition -- 3.1.3 Relation to Other Tasks -- 3.1.4 Challenges -- 3.2 Open-set Recognition: Dataset and Protocol -- 3.2.1 Dataset by Language -- 3.2.2 Metrics -- 3.2.3 Protocols -- 3.3 An OSTR Framework -- 3.3.1 Overall Design and Training Behaviors -- 3.3.2 Testing Behaviors -- 3.4 Modules and Variables -- 3.4.1 Representation Space -- 3.4.2 Label-to-Representation Mapping -- 3.4.3 Sample-to-Representation Mapping -- 3.4.4 Open-set Predictor -- 3.5 Backward Compatibility -- 3.5.1 Part Based -- 3.5.2 Glyph Based -- References -- 4 Open-Set Text Recognition Implementations(I): Label-to-Representation Mapping -- 4.1 Representation Space -- 4.1.1 Gram-based Prototypes -- 4.1.2 Character-Based Prototypes -- 4.1.3 Part-Based Representation -- 4.1.4 Representation Fusing -- 4.2 Label-to-Representation Mapping -- 4.2.1 Side-Information -- 4.2.2 Common Mapping Patterns -- References -- 5 Open-Set Text Recognition Implementations(II): Sample-to-Representation Mapping -- 5.1 Feature Extractor -- 5.1.1 Knowledge-Driven Feature Extractor -- 5.1.2 Data-Driven Feature Extractor -- 5.2 Sampler -- 5.2.1 Gram Based -- 5.2.2 Feature Aggregation -- 5.2.3 Label Aggregation -- 5.3 Linguistic Information Handling -- 5.3.1 Handling Linguistic Information in Close-Environment -- 5.3.2 Handling Linguistic Information in Open-Environment -- References. ; 6 Open-Set Text Recognition Implementations(III): Open-set Predictor -- 6.1 Recognition -- 6.1.1 Simple Distance Function -- 6.1.2 Discrete Attribute Matching -- 6.1.3 Learnable Comparison Function -- 6.2 Rejection of Out-of-Set -- 6.2.1 Centered Approaches -- 6.2.2 Non-Centered Approaches -- 6.3 Cognition -- References -- 7 Open-Set Text Recognition: Case-Studies -- 7.1 OSOCR -- 7.1.1 Framework Implementation Overview -- 7.1.2 Topology-Preserving Transformation Network -- 7.1.3 Label-to-Representation Mapping Module -- 7.1.4 Open-Set Predictor -- 7.1.5 Optimization -- 7.2 Character-Context Decoupling -- 7.2.1 Character-Context Decoupling Framework -- 7.2.2 Decoupled Context Anchor Mechanism -- 7.2.3 blackOpenCCD Network -- 7.2.4 Proof of Theorem 1 -- 7.2.5 Proof of Theorem 2 -- 7.2.6 Proof of Theorem 3 -- 7.3 Performances Overview -- 7.3.1 Open-Set Text Recognition -- 7.3.2 Performance on Other Unseen Languages -- 7.3.3 Standard Close-Set Text Recognition -- References -- 8 Discussions and Future Directions -- 8.1 Applications for OSTR -- 8.2 Discussions on MLLM and OSTR -- 8.3 Future Directions -- References.
• Titolo autorizzato: Open-Set Text Recognition
• ISBN: 981-9703-61-1
• Formato: Materiale a stampa
• Livello bibliografico: Monografia
• Lingua di pubblicazione: Inglese
• Record Nr.: 9910847068803321
• Lo trovi qui: Univ. Federico II
• Serie: SpringerBriefs in Computer Science Series