AbstractIn this paper, we investigate the application of Vehicle-to-Everything (V2X) communication to improve perception performance autonomous vehicles. We present a robust cooperative framework with V2X using novel vision Transformer. Specifically, build holistic attention model, namely V2X-ViT, effectively fuse information across on-road agents (i.e., vehicles and infrastructure). V2X-ViT co...

Visual media permeates and increasingly dominates Western culture. In nearly every area of our lives, at work and leisure, in everything from movies and computer operating systems to television programs and magazine ads, we are bombarded with images and symbols. Even the imaging methods common in modern medical practice, the sonogram, MRI, and CAT scan, are familiar to us independent of direct ...

Previous research has shown that visual perception is affected by sensory information from other modalities. For example, sound can alter the visual intensity or the number of visual objects perceived. However, when touch and vision are combined, vision normally dominates--a phenomenon known as visual capture. Here we report a cross-modal interaction between active touch and vision: The perceiv...

Several studies have shown that adults integrate visual and haptic information (and information from other modalities) in a statistically optimal fashion, weighting each sense according to its reliability [1, 2]. When does this capacity for crossmodal integration develop? Here, we show that prior to 8 years of age, integration of visual and haptic spatial information is far from optimal, with e...