Identifying perceptual thresholds is critical for understanding the mechanisms that underlie signalevolution. Using computer-animated stimuli, we examined visual speed sensitivity in the Jackydragon Amphibolurus muric...Identifying perceptual thresholds is critical for understanding the mechanisms that underlie signalevolution. Using computer-animated stimuli, we examined visual speed sensitivity in the Jackydragon Amphibolurus muricatus, a species that makes extensive use of rapid motor patterns in so-cial communication. First, focal lizards were tested in discrimination trials using random-dot kine-matograms displaying combinations of speed, coherence, and direction. Second, we measuredsubject lizards＇ ability to predict the appearance of a secondary reinforcer （1 of 3 differentcomputer-generated animations of invertebrates： cricket, spider, and mite） based on the directionof movement of a field of drifting dots by following a set of behavioural responses （e.g., orientingresponse, latency to respond） to our virtual stimuli. We found an effect of both speed and coher-ence, as well as an interaction between these 2 factors on the perception of moving stimuli.Overall, our results showed that Jacky dragons have acute sensitivity to high speeds. We then em-ployed an optic flow analysis to match the performance to ecologically relevant motion. Our resultssuggest that the Jacky dragon visual system may have been shaped to detect fast motion. Thispre-existing sensitivity may have constrained the evolution of conspecific displays. In contrast,Jacky dragons may have difficulty in detecting the movement of ambush predators, such as snakesand of some invertebrate prey. Our study also demonstrates the potential of the computer-animated stimuli technique for conducting nonintrusive tests to explore motion range and sensitiv-ity in a visually mediated species.展开更多
由于人左右眼间距的存在,使得同一空间物体在左右眼视网膜上的投影存在位置差异,称之为视差.左右眼视网膜获取的信息最初在初级视皮层(V1区)进行融合,该区域有大量对视差敏感的神经元.关于它们的视差选择特性,目前比较公认的计算模型是...由于人左右眼间距的存在,使得同一空间物体在左右眼视网膜上的投影存在位置差异,称之为视差.左右眼视网膜获取的信息最初在初级视皮层(V1区)进行融合,该区域有大量对视差敏感的神经元.关于它们的视差选择特性,目前比较公认的计算模型是视差能量模型,然而该模型却无法解释V1区神经元对反相关随机点立体图(Anti-correlated random dot stereograms,aRDS)的响应要比对随机点立体图的响应弱这一神经生理学发现.为此,本文提出了一种加权视差能量模型:首先,利用左右眼感受野内的信号差异对神经元的响应能量进行调制,然后再结合神经元之间的相互作用来计算细胞群响应,从而得到图像视差.本文旨在探索基于神经生理学的视差计算方法,主要贡献有:1)加权视差能量模型能够很好地解释V1区神经元对反随机点立体图的响应比随机点立体图响应弱的生理特性;2)加权视差能量模型的视差计算结果精度比现有基于神经生理学的模型更高,甚至高于一些传统的计算机视觉方法.展开更多
文摘Identifying perceptual thresholds is critical for understanding the mechanisms that underlie signalevolution. Using computer-animated stimuli, we examined visual speed sensitivity in the Jackydragon Amphibolurus muricatus, a species that makes extensive use of rapid motor patterns in so-cial communication. First, focal lizards were tested in discrimination trials using random-dot kine-matograms displaying combinations of speed, coherence, and direction. Second, we measuredsubject lizards＇ ability to predict the appearance of a secondary reinforcer （1 of 3 differentcomputer-generated animations of invertebrates： cricket, spider, and mite） based on the directionof movement of a field of drifting dots by following a set of behavioural responses （e.g., orientingresponse, latency to respond） to our virtual stimuli. We found an effect of both speed and coher-ence, as well as an interaction between these 2 factors on the perception of moving stimuli.Overall, our results showed that Jacky dragons have acute sensitivity to high speeds. We then em-ployed an optic flow analysis to match the performance to ecologically relevant motion. Our resultssuggest that the Jacky dragon visual system may have been shaped to detect fast motion. Thispre-existing sensitivity may have constrained the evolution of conspecific displays. In contrast,Jacky dragons may have difficulty in detecting the movement of ambush predators, such as snakesand of some invertebrate prey. Our study also demonstrates the potential of the computer-animated stimuli technique for conducting nonintrusive tests to explore motion range and sensitiv-ity in a visually mediated species.
文摘由于人左右眼间距的存在,使得同一空间物体在左右眼视网膜上的投影存在位置差异,称之为视差.左右眼视网膜获取的信息最初在初级视皮层(V1区)进行融合,该区域有大量对视差敏感的神经元.关于它们的视差选择特性,目前比较公认的计算模型是视差能量模型,然而该模型却无法解释V1区神经元对反相关随机点立体图(Anti-correlated random dot stereograms,aRDS)的响应要比对随机点立体图的响应弱这一神经生理学发现.为此,本文提出了一种加权视差能量模型:首先,利用左右眼感受野内的信号差异对神经元的响应能量进行调制,然后再结合神经元之间的相互作用来计算细胞群响应,从而得到图像视差.本文旨在探索基于神经生理学的视差计算方法,主要贡献有:1)加权视差能量模型能够很好地解释V1区神经元对反随机点立体图的响应比随机点立体图响应弱的生理特性;2)加权视差能量模型的视差计算结果精度比现有基于神经生理学的模型更高,甚至高于一些传统的计算机视觉方法.
