A5048515486- Advanced Image and Video Retrieval Techniques
- Multimodal Machine Learning Applications
- Adversarial Robustness in Machine Learning
- Anomaly Detection Techniques and Applications
- Domain Adaptation and Few-Shot Learning
- Machine Learning in Healthcare
- Advanced Malware Detection Techniques
- Video Analysis and Summarization
- Face recognition and analysis
- Advanced Neural Network Applications
- Advanced Steganography and Watermarking Techniques
Information Technologies Institute
2021-2023
Centre for Research and Technology Hellas
2020-2022
In this paper, we address the problem of global-scale image geolocation, proposing a mixed classification-retrieval scheme. Unlike other methods that strictly tackle as classification or retrieval task, combine two practices in unified solution leveraging advantages each approach with different modules. The first leverages EfficientNet architecture to assign images specific geographic cell robust way. second introduces new residual is trained contrastive learning map input an embedding space...
Enabled by recent improvements in generation methodologies, DeepFakes have become mainstream due to their increasingly better visual quality, the increase easy-to-use tools and rapid dissemination through social media. This fact poses a severe threat our societies with potential erode cohesion influence democracies. To mitigate threat, numerous DeepFake detection schemes been introduced literature but very few provide web service that can be used wild. In this paper, we introduce MeVer...
This paper explores the development of a framework for content-aware user profiling, studying how image producers and consumers can be better understood consequently served through services such as matchmaking friend recommendations. User interests similarities are extracted analyzed on edge employing state art CNN models over images tasks classification, well as, building latent representations from personal media content. A private-by-design approach is employed deployment on-device,...
Artificial intelligence can facilitate the management of large amounts media content and enable organisations to extract valuable insights from their data. Although AI for understanding has made rapid progress over recent years, its deployment in applications professional sectors poses challenges, especially organizations with no expertise. This motivated creation Media Asset Annotation Management platform (MAAM) that employs state-of-the-art deep learning models annotate image video assets....
Enabled by recent improvements in generation methodologies, DeepFakes have become mainstream due to their increasingly better visual quality, the increase easy-to-use tools and rapid dissemination through social media. This fact poses a severe threat our societies with potential erode cohesion influence democracies. To mitigate threat, numerous DeepFake detection schemes been introduced literature but very few provide web service that can be used wild. In this paper, we introduce MeVer...
In this paper, we address the problem of global-scale image geolocation, proposing a mixed classification-retrieval scheme. Unlike other methods that strictly tackle as classification or retrieval task, combine two practices in unified solution leveraging advantages each approach with different modules. The first leverages EfficientNet architecture to assign images specific geographic cell robust way. second introduces new residual is trained contrastive learning map input an embedding space...