le estoy cogiendo un asco tremendo a la FUENTE de las infografías generadas por AI que se usan en Linkedin… no puedo desverla, la detecto al momento, y me genera un rechazo brutal
a las infografías también, obvio, porque son cosas muy fáciles de hacer a manita sin tener que pasar por genAI, pero la fuente en concreto me tiltea
Eight photos that make us question what we see
"The effect that scares me most is not that we'll be fooled by fake photos but that we'll ignore the real ones" – how photographers are dealing with shifting perceptions of reality.
3 current photo exhibitions mentioned in the article, in Norwich, Amsterdam, & Maastricht
#Photography #GenerativeAI #GeneratedAI #AIPhoto
https://www.bbc.com/culture/article/20240711-eight-photos-that-make-us-question-what-we-see
At #ESWC2024 we presented our poster "Gotta Catch’em All: From Data Silos to a Knowledge Graph" with the @nfdi4culture data harvesting pipeline which is supposed to harvest, clean, map & integrate data into the NFDI4Culture-KG. Joint work of @sashabruns @tabea @jalle @epoz @lysander07 Linnea Söhn & Torsten Schrade
paper: https://zenodo.org/records/11505790
#knowledgegraphs #nfdirocks #semanticweb #poster #generatedAI #pokemon @fiz_karlsruhe
Diverse research questions, perspectives, standards and formats across culture subject areas have led to the emergence of numerous data silos. NFDI4Culture seeks to overcome this by building a unified KG, facilitating enhanced discoverability and interoperability of distributed and heterogeneous research data. This paper outlines a pipeline for accessing and harvesting cultural heritage meta data from legacy repositories, it discusses the development of a lightweight ontology to facilitate interoperability.
#AI #GenerativeAI #GeneratedAI: "The advent of generative AI images has completely disrupted the art world. Distinguishing AI generated images from human art is a challenging problem whose impact is growing over time. A failure to address this problem allows bad actors to defraud individuals paying a premium for human art and companies whose stated policies forbid AI imagery. It is also critical for content owners to establish copyright, and for model trainers interested in curating training data in order to avoid potential model collapse.
There are several different approaches to distinguishing human art from AI images, including classifiers trained by supervised learning, research tools targeting diffusion models, and identification by professional artists using their knowledge of artistic techniques. In this paper, we seek to understand how well these approaches can perform against today's modern generative models in both benign and adversarial settings. We curate real human art across 7 styles, generate matching images from 5 generative models, and apply 8 detectors (5 automated detectors and 3 different human groups including 180 crowdworkers, 4000+ professional artists, and 13 expert artists experienced at detecting AI). Both Hive and expert artists do very well, but make mistakes in different ways (Hive is weaker against adversarial perturbations while Expert artists produce higher false positives). We believe these weaknesses will remain as models continue to evolve, and use our data to demonstrate why a combined team of human and automated detectors provides the best combination of accuracy and robustness." https://arxiv.org/abs/2402.03214
The advent of generative AI images has completely disrupted the art world. Distinguishing AI generated images from human art is a challenging problem whose impact is growing over time. A failure to address this problem allows bad actors to defraud individuals paying a premium for human art and companies whose stated policies forbid AI imagery. It is also critical for content owners to establish copyright, and for model trainers interested in curating training data in order to avoid potential model collapse. There are several different approaches to distinguishing human art from AI images, including classifiers trained by supervised learning, research tools targeting diffusion models, and identification by professional artists using their knowledge of artistic techniques. In this paper, we seek to understand how well these approaches can perform against today's modern generative models in both benign and adversarial settings. We curate real human art across 7 styles, generate matching images from 5 generative models, and apply 8 detectors (5 automated detectors and 3 different human groups including 180 crowdworkers, 4000+ professional artists, and 13 expert artists experienced at detecting AI). Both Hive and expert artists do very well, but make mistakes in different ways (Hive is weaker against adversarial perturbations while Expert artists produce higher false positives). We believe these weaknesses will remain as models continue to evolve, and use our data to demonstrate why a combined team of human and automated detectors provides the best combination of accuracy and robustness.
elenadotnet
Werner Kratz
sadele2
FIZ ISE Research Group
Miguel Afonso Caetano
LukaszD