As indigenous fabrics gain renewed visibility in contemporary fashion and visual art, artist and cultural advocate Goodluck ...

Researchers have identified a new type of visual cell in deep-sea fish larvae that challenges a century of knowledge about ...

Despite the display’s mass appeal, “Cunning Folk” defies expectations. Historically, most media themed around witches has ...

Think about the last professional video you watched—a commercial, a music video, a film trailer. Beyond the cinematography and editing, you probably noticed something else: the effects and transitions ...

Abstract: Prompt tuning is a valuable technique for adapting visual language models (VLMs) to different downstream tasks, such as domain generalization and learning from a few examples. Previous ...

A new dance-centric work from longtime Seattle interdisciplinary artists probes the balance of masculine and feminine, nature ...

Material objects communicate in a way that’s fundamentally different from everything else,” the artist says. “They tell you so much—not just about the maker, but about the materials, ...

Abstract: Knowledge is an abstraction of factual principles of the physical world. Large foundation models encapsulate extensive multimodal knowledge into the parameters and thus invoke machine ...

The agent acquires a vocabulary of neuro-symbolic concepts for objects, relations, and actions, represented through a combination of symbolic programs and neural networks. These concepts are grounded ...

In this paper, we focus on challenging task driven object detection, which is practical in the real world yet underexplored. To bridge the gap between abstract task requirements and objects in the ...

Imagine an assembly line where each tool monitors how it’s used; errors are detected before they occur; and digital work instructions adjust in real time based on live feedback. That concept is now ...

Distilling knowledge from the pre-trained teacher models helps to learn a small student model that generalizes better. While existing works mostly focus on distilling knowledge within the same ...