Generative AI in Art and Vision

The field of generative AI is rapidly advancing, with a focus on creating innovative and interactive digital content. Recent developments have enabled the automated generation of high-quality 3D environments, human animations, and synthetic data for various applications. The integration of expert-driven motion encoding, prompt-guided avatar generation, and human-aware background synthesis has led to the creation of highly varied and lifelike human sequences. Additionally, the use of generative AI in art and cultural heritage has opened up new research perspectives, allowing for the production of self-explaining cultural artifacts and the exploration of digital images in relation to their malleability and contemporary interpretation. Noteworthy papers include: Gen4D, which introduces a fully automated pipeline for generating diverse and photorealistic 4D human animations, and EmbodiedGen, which presents a foundational platform for interactive 3D world generation, enabling the scalable generation of high-quality, controllable, and photorealistic 3D assets. Overall, the field is moving towards more realistic, interactive, and accessible digital content generation, with potential applications in industries such as gaming, virtual reality, and cinema.

Sources

Speaking images. A novel framework for the automated self-description of artworks

Gen4D: Synthesizing Humans and Scenes in the Wild

AI-powered Contextual 3D Environment Generation: A Systematic Review

Can Artificial Intelligence Write Like Borges? An Evaluation Protocol for Spanish Microfiction

An Effective End-to-End Solution for Multimodal Action Recognition

Synthetic Human Action Video Data Generation with Pose Transfer

Using Sign Language Production as Data Augmentation to enhance Sign Language Translation

A Survey of Automatic Evaluation Methods on Text, Visual and Speech Generations

EmbodiedGen: Towards a Generative 3D World Engine for Embodied Intelligence

Built with on top of