A5008204423- Aesthetic Perception and Analysis
- Visual Attention and Saliency Detection
- Conservation Techniques and Studies
- Advanced Image and Video Retrieval Techniques
- Multimodal Machine Learning Applications
- Advanced Data Storage Technologies
- Scientific Computing and Data Management
- Distributed and Parallel Computing Systems
- Human Pose and Action Recognition
- Art History and Market Analysis
- Digital Media and Visual Art
- 3D Surveying and Cultural Heritage
- Cultural Heritage Management and Preservation
- Ethics and Social Impacts of AI
- Museums and Cultural Heritage
- Computational and Text Analysis Methods
- Generative Adversarial Networks and Image Synthesis
- Image Retrieval and Classification Techniques
- Cultural Heritage Materials Analysis
- FinTech, Crowdfunding, Digital Finance
- Image Enhancement Techniques
- Hand Gesture Recognition Systems
- Video Analysis and Summarization
University of Zurich
2021-2023
Rudjer Boskovic Institute
2016-2022
Durham University
2021-2022
University of Zagreb
2013
Technologies related to artificial intelligence (AI) have a strong impact on the changes of research and creative practices in visual arts. The growing number initiatives applications that emerge intersection AI art motivates us examine discuss explorative potentials technologies context art. This article provides an integrated review two facets art: (1) is used for analysis employed digitized artwork collections, or (2) purposes generating novel artworks. In AI-related understanding, we...
With the emergence of large digitized fine art collections and successful performance deep learning techniques, new research prospects unfold in intersection artificial intelligence art. In order to explore applicability techniques understanding images beyond object recognition classification, we employ convolutional neural networks (CNN) predict scores related three subjective aspects human perception: aesthetic evaluation image, sentiment evoked by memorability image. For each concept,...
This paper describes the achievements of H2020 project INDIGO-DataCloud. The has provided e-infrastructures with tools, applications and cloud framework enhancements to manage demanding requirements scientific communities, either locally or through enhanced interfaces. middleware developed allows federate hybrid resources, easily write, port run cloud. In particular, we have extended existing PaaS (Platform as a Service) solutions, allowing public private e-infrastructures, including those...
Hands represent an important aspect of pictorial narration but have rarely been addressed as object study in art history and digital humanities. Although hand gestures play a significant role conveying emotions, narratives, cultural symbolism the context visual art, comprehensive terminology for classification depicted poses is still lacking. In this article, we present process creating new annotated dataset poses. The based on collection European early modern paintings, from which hands are...
To automatically generate accurate and meaningful textual descriptions of images is an ongoing research challenge. Recently, a lot progress has been made by adopting multimodal deep learning approaches for integrating vision language. However, the task developing image captioning models most commonly addressed using datasets natural images, while not many contributions have in domain artwork images. One main reasons that lack large-scale art adequate image-text pairs. Another reason fact...
Extensive digitization efforts in the recent years have led to a large increase of digitized and online available fine-art collections. With artworks, we aim preserve all those valuable evidences various human creative expressions, as well make them broader audience. The digitalization process artworks should not constrain only fulfilling purpose preservation, but also serve starting point for exploring this type data novel way, which is made possible with rise new achievements computer...
Technologies related to artificial intelligence (AI) have a strong impact on the changes of research and creative practices in visual arts. The growing number initiatives applications that emerge intersection AI art, motivates us examine discuss explorative potentials technologies context art. This paper provides an integrated review two facets art: 1) is used for art analysis employed digitized artwork collections; 2) purposes generating novel artworks. In AI-related understanding, we...
In the framework of H2020 INDIGO-DataCloud project, we have implemented an advanced solution for automatic deployment digital data repositories based on Invenio, library developed by CERN. Exploiting cutting-edge technologies, such as Docker and Apache Mesos, standard specifications to describe application architectures TOSCA, are able provide a service that simplifies process creating managing various assets using cloud resources. An Invenio-based repository consists set services (e.g....
Text-to-image (TTI) systems, particularly those utilizing open-source frameworks, have become increasingly prevalent in the production of Artificial Intelligence (AI)-generated visuals. While existing literature has explored various problematic aspects TTI technologies, such as bias generated content, intellectual property concerns, and reinforcement harmful stereotypes, frameworks not yet been systematically examined from a cultural perspective. This study addresses this gap by analyzing...
Recent critiques of Artificial-intelligence (AI)-generated visual content highlight concerns about the erosion artistic originality, as these systems often replicate patterns from their training datasets, leading to significant uniformity and reduced diversity. Our research adopts a novel approach by focusing on user behavior during interactions with Text-to-Image models. Instead solely analyzing data patterns, we examine how users' tendencies create original prompts or rely common templates...
Inspired by the successful performance of Convolutional Neural Networks (CNN) in automatically predicting complex image properties such as memorability, this work we explore their transferability to domain art images. We employ a CNN model trained predict memorability scores natural images artworks belonging different genres and styles. Our experiments show that nude painting portrait are most memorable genres, while landscape marine least memorable. Regarding style, abstract styles tend be...
The paper discusses the potential of large vision-language models as objects interest for empirical cultural studies. Focusing on comparative analysis outputs from two popular text-to-image synthesis models, DALL-E 2 and Stable Diffusion, tries to tackle pros cons striving towards culturally agnostic vs. specific AI models. several examples memorization bias in generated which showcase trade-off between risk mitigation specificity, well overall impossibility developing
Image captioning implies automatically generating textual descriptions of images based only on the visual input. Although this has been an extensively addressed research topic in recent years, not many contributions have made domain art historical data. In particular context, task image is confronted with various challenges such as lack large-scale datasets image-text pairs, complexity meaning associated describing artworks and need for expert-level annotations. This work aims to address...