作者: Jasmine M. Yakubowski , Glyn A. McMillan , John R. Gray
DOI: 10.14814/PHY2.12801
关键词: Looming 、 Stimulus (physiology) 、 Computer vision 、 Neuron 、 Artificial intelligence 、 Detector 、 Visual perception 、 Collision 、 Motion detection 、 Physics 、 Visual motion
摘要: Stimulus complexity affects the response of looming sensitive neurons in a variety animal taxa. The Lobula Giant Movement Detector/Descending Contralateral Detector (LGMD/DCMD) pathway is well-characterized locust visual system. It responds to simple objects approaching on direct collision course (i.e., looming) as well complex motion defined by changes stimulus velocity, trajectory, and transitions, all which are affected presence or absence background motion. In this study, we focused DCMD responses transitioning away from course, emulates successful avoidance behavior. We presented each 20 locusts with sequence three-dimensional stimuli simple, scattered, progressive flow field backgrounds while simultaneously recording activity extracellularly. were generally characteristic irrespective background. However, changing affected, peak firing rates, time, caused rise fall phases. object also varied azimuthal approach angle dynamics edge expansion. These data fit an existing correlational model that relates expansion properties rate modulation during trajectory changes.