A new arXiv paper tests whether multimodal LLMs can judge visual creativity without examples or fine-tuning. Across AI images and hand-drawn sketches, models aligned with human ratings, while their reasoning exposed what they weighed.
https://t.co/exVwpvH7Xj
A new arXiv paper introduces the Human Creativity Benchmark for evaluating AI in creative work. Across 15,000 professional judgments, it separates shared quality signals from genuine taste differences instead of collapsing both into one score.
https://t.co/VL6AYcQHqG
A new arXiv paper introduces AnimaSpark, a feed-forward method for animating arbitrary 3D objects from static assets. It reports better text-motion alignment, motion quality, and efficiency than prior category-agnostic animation methods.
https://t.co/tAq8QNDrj8
A new arXiv paper reports on students building AI-enabled solutions for cultural heritage preservation and sustainable development. It proposes a culturally-aware AI education framework linking community learning, design, and computation.
https://t.co/9zv7mJZK35
A new paper builds a multimodal knowledge-graph system for Chinese shadow play narratives. It combines vision, text, pose, GANs, and blockchain to generate traceable stories with better user retention and cultural-symbol consistency.
https://t.co/BmnDYGfH5g
A new Frontiers paper introduces an AI-directed PCG framework for personalized XR in an industrial-heritage museum. Visitor profiles from telemetry aligned with self-reports, and heritage perception increased after the session.
https://t.co/bHRAme4UKy
A new paper studies Stable Diffusion XL in digital art classrooms using Gulf cultural heritage prompts. It found recurring errors in attire, historical period, and body forms, making verification literacy a core design skill.
https://t.co/2R2hkdpFQT
A new Frontiers paper introduces HybridWeaveNet for recognizing Indian handloom heritage fabrics. It combines EfficientNetV2 with dual attention, reaching 91% performance and using Grad-CAM to show focus on culturally related motifs.
https://t.co/XTZlD6ZGpl
A new Frontiers paper studied continued use of an intangible-cultural-heritage VR game about the Qinhuai Lantern Festival. It found sensory, emotional, and cultural experience all shape users’ attitudes and intention to keep playing.
https://t.co/3kYiVzx6DH
A European Commission call funds AI integration in cultural and creative sectors through Horizon Europe. It prioritizes practical adoption: tools, pilots, guidance, and training for cultural work.
https://t.co/yI7KrvNhjZ
A new Scientific Reports paper proposes DTH-AI, a stakeholder-based framework for AI-assisted restoration and MR digital heritage. It maps restoration, heritage characterization, digitization, and design into a sustainability pathway.
https://t.co/J7hS5FaT5l
Arup announced DATALAND will open June 20 as a Museum of AI Arts in downtown LA. The project turns generative, multi-sensory artworks into a dedicated museum environment.
https://t.co/KIaNmtshJG
A new paper studied how Chinese visual arts students use AI tools in art education. It found needs differ by gender, degree level, and institution type, arguing AI teaching strategies should be tailored, not one-size-fits-all.
https://t.co/VRNmGWAkny
A new paper introduces SmartTracer, an AI-assisted tool for extracting editable strokes from complex line art. It lets artists guide or refine the result, recovering cleaner stroke masks than existing methods and commercial software.
https://t.co/YaSU47fXAJ
Neue Nationalgalerie presented Beeple’s Regular Animals, an AI-and-robotics installation. Autonomous robots captured the gallery, then AI turned views into printed images shaped by each figure’s worldview.
https://t.co/0Al6PJn0as
A new paper compared audiobook excerpts read by AI-generated voices and professional human narrators. Listeners rated them almost the same, suggesting synthetic voices can now function as viable substitutes for audiobook narration.
https://t.co/rcUs3wTM3e
A new paper uses CAM heatmaps to guide neural super-sampling for intangible cultural heritage rendering. It directs compute to culturally important image regions, making immersive visuals clearer and more interpretable.
https://t.co/jf6nVh45D9
AI in cultural heritage is moving from digitization to interpretation.
New paper uses GANs + human feedback to translate heritage patterns into design. The test: can AI extend tradition without flattening memory?
https://t.co/QVTPWI4SRG