Corinna Reinhardt
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Corinna Reinhardt has a total of 7 co-authors. Here are the top co-authors by number of shared publications:
- Torsten Bendschus
Institut für Klassische Archäologie, FAU, Erlangen, Germany
2 shared publications - Andreas Maier
Friedrich-Alexander-Universität Erlangen-Nürnberg, Pattern Recognition Lab, Erlangen, Germany
1 shared publication - Angel Villar-Corrales
Autonomous Intelligent Systems, University of Bonn, Bonn, North Rhine-Westphalia, Germany
1 shared publication - Peter Bell
Institut für Kunstgeschichte Philipps-Universität Marburg Germany
1 shared publication - Prathmesh Madhu
Pattern Recognition Lab Friedrich-Alexander-Universität Erlangen Germany
1 shared publication - Ronak Kosti
Friedrich-Alexander-Universität, Erlangen-Nürnberg, Germany
1 shared publication
Latest Publications
- Top Cited PapersOpen Access
- 24 December 2022
- Published by Association for Computing Machinery (ACM) in Journal on Computing and Cultural Heritage
- Vol. 16 (1) , 1-17
Enhancing Human Pose Estimation in Ancient Vase Paintings via Perceptually-grounded Style Transfer Learning
Human pose estimation (HPE) is a central part of understanding the visual narration and body movements of characters depicted in artwork collections, such as Greek vase paintings. Unfortunately, existing HPE methods do not generalise well across domains resulting in poorly recognised poses. Therefore, we propose a two step approach: (1) adapting a dataset of natural images of known person and pose annotations to the style of Greek vase paintings by means of image style-transfer. We introduce a perceptually-grounded style transfer training to enforce perceptual consistency. Then, we fine-tune the base model with this newly created dataset. We show that using style-transfer learning significantly improves the SOTA performance on unlabelled data by more than 6% mean average precision (mAP) as well as mean average recall (mAR) . (2) To improve the already strong results further, we created a small dataset (ClassArch) consisting of ancient Greek vase paintings from the 6–5th century BCE with person and pose annotations. We show that fine-tuning on this data with a style-transferred model improves the performance further. In a thorough ablation study, we give a targeted analysis of the influence of style intensities, revealing that the model learns generic domain styles. Additionally, we provide a pose-based image retrieval to demonstrate the effectiveness of our method. The code and pretrained models can be found at https://github.com/angelvillar96/STLPose .
- 24 December 2022
- Published by Association for Computing Machinery (ACM) in Journal on Computing and Cultural Heritage
- Vol. 16 (1) , 1-17
Top Cited PapersOpen AccessEnhancing Human Pose Estimation in Ancient Vase Paintings via Perceptually-grounded Style Transfer Learning
Human pose estimation (HPE) is a central part of understanding the visual narration and body movements of characters depicted in artwork collections, such as Greek vase paintings. Unfortunately, existing HPE methods do not generalise well across domains resulting in poorly recognised poses. Therefore, we propose a two step approach: (1) adapting a dataset of natural images of known person and pose annotations to the style of Greek vase paintings by means of image style-transfer. We introduce a perceptually-grounded style transfer training to enforce perceptual consistency. Then, we fine-tune the base model with this newly created dataset. We show that using style-transfer learning significantly improves the SOTA performance on unlabelled data by more than 6% mean average precision (mAP) as well as mean average recall (mAR) . (2) To improve the already strong results further, we created a small dataset (ClassArch) consisting of ancient Greek vase paintings from the 6–5th century BCE with person and pose annotations. We show that fine-tuning on this data with a style-transferred model improves the performance further. In a thorough ablation study, we give a targeted analysis of the influence of style intensities, revealing that the model learns generic domain styles. Additionally, we provide a pose-based image retrieval to demonstrate the effectiveness of our method. The code and pretrained models can be found at https://github.com/angelvillar96/STLPose .
- 1 November 2022
- Published by Cambridge University Press (CUP) in The Journal of Hellenic Studies
- Vol. 142, 462-464
(A.) PETSALIS-DIOMIDIS with (E.) HALL (eds) The Classical Vase Transformed: Consumption, Reproduction, and Class in Eighteenth and Nineteenth-Century Britain. (Bulletin of the Institute of Classical Studies; 63-1, 2020.) London: Institute of Classical Studies, School of Advanced Study, University of London, 2020. Pp. 142, illus. ISSN 0076-0730.
- 1 November 2022
- Published by Cambridge University Press (CUP) in The Journal of Hellenic Studies
- Vol. 142, 462-464
(A.) PETSALIS-DIOMIDIS with (E.) HALL (eds) The Classical Vase Transformed: Consumption, Reproduction, and Class in Eighteenth and Nineteenth-Century Britain. (Bulletin of the Institute of Classical Studies; 63-1, 2020.) London: Institute of Classical Studies, School of Advanced Study, University of London, 2020. Pp. 142, illus. ISSN 0076-0730.
- Open Access
- 13 August 2021
- Published by Universitatsbibliothek Kiel
- p. 117-128
Simulating Digital Classics in Classical Teaching
In the 21st century, Classical Archaeology is making more and more use of digital tools and methods. This tendency towards a future field of “Digital Classics” requires participation not only as users, but also as developers. For this reason, the required qualification profile for a student of Classical Archaeology is changing and academic teaching at universities is confronted with new challenges. Our presentation tackles this issue by suggesting a new teaching concept that focuses especially on the pivotal skills that are needed to use and develop digital methods within an interdisciplinary team. It is based on the didactic model of a simulation game. This simulation is attached to a (real) interdisciplinary research project. In this way it offers the possibility of a structured process model and challenges the participants’ skills of interaction and complex decision-making. The result is a realistic environment whose demands, means and conditions of action support the assessment and evaluation of academic expectations in multidisciplinary professional situations
- 13 August 2021
- Published by Universitatsbibliothek Kiel
- p. 117-128
Open AccessSimulating Digital Classics in Classical Teaching
In the 21st century, Classical Archaeology is making more and more use of digital tools and methods. This tendency towards a future field of “Digital Classics” requires participation not only as users, but also as developers. For this reason, the required qualification profile for a student of Classical Archaeology is changing and academic teaching at universities is confronted with new challenges. Our presentation tackles this issue by suggesting a new teaching concept that focuses especially on the pivotal skills that are needed to use and develop digital methods within an interdisciplinary team. It is based on the didactic model of a simulation game. This simulation is attached to a (real) interdisciplinary research project. In this way it offers the possibility of a structured process model and challenges the participants’ skills of interaction and complex decision-making. The result is a realistic environment whose demands, means and conditions of action support the assessment and evaluation of academic expectations in multidisciplinary professional situations