Active Visual Depth Perception by Looming

Looming 的主动视觉深度感知

• 批准号: 9975048
• 负责人: Terrence Sejnowski
• 金额: $ 12.91万
• 依托单位: The Salk Institute for Biological Studies
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 1999
• 资助国家: 美国
• 起止时间: 1999-08-15 至 2001-07-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9975048&HistoricalAwards=false
• 关键词: Active; Visual; Depth; Perception; Looming

This proposal is a project to measure and model looming responses of cellsin the optic ganglia (primarily the lobula) of the night-flying hawk moth,Manduca sexta. When the moth hovers in front of a flower to feed, windcauses the flower to move in the moth's visual field. The experimentsproposed are of two kinds: First, responses to a wide variety of visualstimuli will be recorded intracellularly as well as extracellularly fromlooming neurons in order to test if the trigger for the looming responsesare stimulus changes of luminance, changes of perimeter, or changes due tovisual flow fields. Second, computer models of these neuronal responseswill be constructed and tested with equivalent 'stimuli'. It is planned totest the models with real images by feeding into the models videos of e.g.foliage or flowers wiggling in the wind. One very practical future use ofsuch computer models may be in machine (robot) vision and in neuromorphicchips.

这项建议是一个项目，以测量和模拟细胞的隐约反应的视神经节（主要是小叶）的夜间飞行鹰蛾，天蛾。当蛾盘旋在花前觅食时，风会使花在蛾的视野中移动。实验分为两类：第一，从隐现神经元细胞内和细胞外记录对各种视觉刺激的反应，以测试隐现反应的触发因素是亮度的刺激变化、周长的变化还是视觉流场的变化。其次，这些神经元反应的计算机模型将被构建并用等效的“刺激”进行测试。计划通过向模型中输入树叶或花朵在风中摆动的视频，用真实的图像测试模型。这种计算机模型的一个非常实际的未来用途可能是机器（机器人）视觉和神经形态芯片。