AI News

Researchers from Capital Normal University and the School of Artificial Intelligence at Beijing University of Posts and Telecommunications have developed RealNet, a new feature reconstruction framework for industrial image anomaly detection. This approach addresses ongoing issues with generating diverse, realistic anomalies that align with natural distributions, as well as challenges around feature redundancy and pre-training…

A research team from the Korea Advanced Institute of Science and Technology (KAIST) has contributed to the field of machine interpretation and interaction which amalgamates AI’s language understanding and visual perception, with the development of MoAI. The model utilizes auxiliary visual information from specialized computer vision (CV) models, which provides a more nuanced understanding of…

Visual Language Models (VLMs), which are powerful tools for processing visual and textual data, can face difficulties due to limited data availability. Recent research developments have shown that pre-training these models on larger image-text datasets can enhance their performance in downstream tasks. However, creating these datasets can be challenging because of paired data scarcity, high…

Image generation from textual descriptions has revolutionized the way technology intersects with creativity. A domain that has garnered interest is subject-driven image generation. Its potential lies in creating personalized images of specific subjects from a minimal set of examples. Yet, the inability to fully capture and depict detailed attributes of a given subject within its…

In the field of 3D generative AI, a new dimension has emerged whereby 3D reconstruction can occur from limited views. Propelled by large-scale 3D datasets and advances in generative model topologies, research has been spearheaded into using 2D diffusion models to create 3D objects from input texts or photos. This is primarily to address the…

The creation of lifelike images, videos, and sounds using artificial intelligence (AI) has significantly progressed recently. However, most of these developments have been focused on single modalities, ignoring the inherent multimodal nature of our world. In addressing this, researchers have introduced a novel optimization-based framework designed to seamlessly integrate visual and audio content creation. By…

Modern vision-language models (VLMs) have made significant progress in providing solutions for multimodal tasks by merging the reasoning abilities of large language models (LLMs) and visual encoders like ViT. Nevertheless, despite their impressive performance in tasks involving entire images, these models often struggle with the fine-grained region grounding, inter-object spatial relations, and compositional reasoning. They…