NSTL国家科技图书文献中心

1. KaiRacters: Character-Level-Based Writer Retrieval for Greek Papyri NSTL国家科技图书文献中心

Marco Peer | Robert Sablatnig ... - 《Pattern Recognition,Part XIX》 - International Conference on Pattern Recognition - 2025, - 73～88 - 共16页

摘要：This paper presents a character-based approach for enhancing writer retrieval performance in the context of Greek papyri. Our contribution lies in introducing character-level annotations for frequentl...

关键词： Greek papyri | Writer retrieval | Historical documents | Document analysis

2. Advancing Handwritten Text Detection by Synthetic Text NSTL国家科技图书文献中心

Markus Muth | Marco Peer ... - 《Pattern Recognition,Part XIX》 - International Conference on Pattern Recognition - 2025, - 122～136 - 共15页

摘要：In this paper, the usability of synthetic handwritten text to improve machine learning models is examined for the domain of handwritten text detection. We generate synthetic handwritten text by using ...

关键词： Handwritten text detection | Synthetic data | Deep learning

3. Fusing Forces: Deep-Human-Guided Refinement of Segmentation Masks NSTL国家科技图书文献中心

Rafael Sterzinger | Christian Stippel ... - 《Pattern Recognition,Part XII》 - International Conference on Pattern Recognition - 2025, - 154～169 - 共16页

摘要：Etruscan mirrors constitute a significant category in Etruscan art, characterized by elaborate figurative illustrations featured on their backside. A laborious and costly aspect of their analysis and ...

关键词： Binarization | Interactive segmentation | Human-in-the-Loop | Etruscan art | Cultural heritage

4. Classical Photometric Stereo in Point Lighting Environments: Error Analysis and Mitigation NSTL国家科技图书文献中心

Simon Brenner | Robert Sablatnig - 《2024 International Conference on 3D Vision: 3DV 2024, Davos, Switzerland, 18-21 March 2024, [v.1]》 - International Conference on 3D Vision - 2024, - 581～590 - 共10页

摘要：While classical photometric stereo assumes parallel lighting and thus allows for efficient solutions, this assumption is rarely met in practice. In acquisition setups using point-like light sources su...

关键词： Reflectivity | Upper bound | Monte Carlo methods | Uncertainty | Three-dimensional displays | Error analysis | Lighting

5. Maximizing Data Efficiency of HTR Models by Synthetic Text NSTL国家科技图书文献中心

Markus Muth | Marco Peer ... - 《Document Analysis Systems: 16th IAPR International Workshop, DAS 2024, Athens, Greece, August 30-31, 2024, Proceedings》 - IAPR International Workshop on Document Analysis Systems - 2024, - 295～311 - 共17页

摘要：The usability of synthetic handwritten text to improve machine learning models is assessed for the domain of HTR. Synthetic handwritten text is generated using an existing model based on a GAN. The ou...

关键词： Synthetic data | Handwritten text recognition | Synthetic text

6. Drawing the Line: Deep Segmentation for Extracting Art from Ancient Etruscan Mirrors NSTL国家科技图书文献中心

Rafael Sterzinger | Simon Brenner ... - 《Document Analysis and Recognition - ICDAR 2024: 18th International Conference, Athens, Greece, August 30-September 4, 2024, Proceedings, Part III》 - International Conference on Document Analysis and Recognition - 2024, - 39～56 - 共18页

摘要：Etruscan mirrors constitute a significant category within Etruscan art and, therefore, undergo systematic examinations to obtain insights into ancient times. A crucial aspect of their analysis involve...

关键词： Binarization | Photometric stereo | Image segmentation | Limited data | Etruscan art | Cultural heritage

7. SAGHOG: Self-supervised Autoencoder for Generating HOG Features for Writer Retrieval NSTL国家科技图书文献中心

Marco Peer | Florian Kleber ... - 《Document Analysis and Recognition - ICDAR 2024: 18th International Conference, Athens, Greece, August 30-September 4, 2024, Proceedings, Part II》 - International Conference on Document Analysis and Recognition - 2024, - 121～138 - 共18页

摘要：This paper introduces Saghog, a self-supervised pretraining strategy for writer retrieval using HOG features of the binarized input image. Our preprocessing involves the application of the Segment Any...

关键词： Writer retrieval | Self-Supervised learning | Masked autoencoder | Document analysis

8. ECSIC: Epipolar Cross Attention for Stereo Image Compression NSTL国家科技图书文献中心

Matthias W?dlinger | Jan Kotera ... - 《2024 IEEE/CVF Winter Conference on Applications of Computer Vision: WACV 2024, Waikoloa, Hawaii, USA, 3-8 January 2024, [v.11]》 - IEEE/CVF Winter Conference on Applications of Computer Vision - 2024, - 3424～3433 - 共10页

摘要：In this paper, we present ECSIC, a novel learned method for stereo image compression. Our proposed method compresses the left and right images in a joint manner by exploiting the mutual information be...

关键词： Computer vision | Image coding | Computational modeling | Estimation | Benchmark testing | Entropy | Decoding

9. Towards Writer Retrieval for Historical Datasets NSTL国家科技图书文献中心

Marco Peer | Florian Kleber ... - 《Document Analysis and Recognition: 17th International Conference, San Jose, CA, USA, August 21-26, 2023, Proceedings, Part I》 - International Conference on Document Analysis and Recognition - 2023, - 411～427 - 共17页

摘要：This paper presents an unsupervised approach for writer retrieval based on clustering SIFT descriptors detected at keypoint locations resulting in pseudo-cluster labels. With those cluster labels, a r...

关键词： Writer retrieval | NetVLAD | Reranking | Document analysis

10. Color-Light Multi Cascade Network for Single Image Depth Prediction on One Perspective Artifact Images NSTL国家科技图书文献中心

Aufaclav Zatu Kusuma... | Simon Brenner ... - 《17th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications: VISIGRAPP 2022, Online, 6-8 February 2022, v.4,p.2. VISAPP》 - International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - 2022, - 909～916 - 共8页

摘要：Different color material and extreme lighting change pose a problem for single image depth prediction on archeological artifacts. These conditions can lead to misprediction on the surface of the foreg...

摘要：This paper presents a character-based approach for enhancing writer retrieval performance in the context of Greek papyri. Our contribution lies in introducing character-level annotations for frequentl...

关键词： Greek papyri | Writer retrieval | Historical documents | Document analysis

摘要：In this paper, the usability of synthetic handwritten text to improve machine learning models is examined for the domain of handwritten text detection. We generate synthetic handwritten text by using ...

关键词： Handwritten text detection | Synthetic data | Deep learning

摘要：Etruscan mirrors constitute a significant category in Etruscan art, characterized by elaborate figurative illustrations featured on their backside. A laborious and costly aspect of their analysis and ...

关键词： Binarization | Interactive segmentation | Human-in-the-Loop | Etruscan art | Cultural heritage

摘要：While classical photometric stereo assumes parallel lighting and thus allows for efficient solutions, this assumption is rarely met in practice. In acquisition setups using point-like light sources su...

关键词： Reflectivity | Upper bound | Monte Carlo methods | Uncertainty | Three-dimensional displays | Error analysis | Lighting

摘要：The usability of synthetic handwritten text to improve machine learning models is assessed for the domain of HTR. Synthetic handwritten text is generated using an existing model based on a GAN. The ou...

关键词： Synthetic data | Handwritten text recognition | Synthetic text

摘要：Etruscan mirrors constitute a significant category within Etruscan art and, therefore, undergo systematic examinations to obtain insights into ancient times. A crucial aspect of their analysis involve...

关键词： Binarization | Photometric stereo | Image segmentation | Limited data | Etruscan art | Cultural heritage

摘要：This paper introduces Saghog, a self-supervised pretraining strategy for writer retrieval using HOG features of the binarized input image. Our preprocessing involves the application of the Segment Any...

关键词： Writer retrieval | Self-Supervised learning | Masked autoencoder | Document analysis

摘要：In this paper, we present ECSIC, a novel learned method for stereo image compression. Our proposed method compresses the left and right images in a joint manner by exploiting the mutual information be...

关键词： Computer vision | Image coding | Computational modeling | Estimation | Benchmark testing | Entropy | Decoding

摘要：This paper presents an unsupervised approach for writer retrieval based on clustering SIFT descriptors detected at keypoint locations resulting in pseudo-cluster labels. With those cluster labels, a r...

关键词： Writer retrieval | NetVLAD | Reranking | Document analysis

摘要：Different color material and extreme lighting change pose a problem for single image depth prediction on archeological artifacts. These conditions can lead to misprediction on the surface of the foreg...

关键词： Single image | Depth prediction | Color-light | Multi cascade | One-side perspective | State-of-the-Art | Roman coins | Temple relief

4008-161-200 800-990-8900

国家科技图书文献中心