Abstract:
Motion perception is a fundamental sensory task that plays a critical evolutionary role. In vision, motion processing is classically described using a motion energy model with spatiotemporally nonseparable filters suited for capturing the smooth continuous changes in spatial position over time afforded by moving objects. However, it is still not clear whether the filters underlying auditory motion discrimination are also continuous motion detectors or infer motion from comparing discrete sound locations over time (spatiotemporally separable). We used a psychophysical reverse correlation paradigm, where participants discriminated the direction of a motion signal in the presence of spatiotemporal noise, to determine whether the filters underlying auditory motion discrimination were spatiotemporally separable or nonseparable. We then examined whether these auditory motion filters were altered as a result of early blindness. We found that both sighted and early blind individuals have separable filters. However, early blind individuals show increased sensitivity to auditory motion, with reduced susceptibility to noise and filters that were more accurate in detecting motion onsets/offsets. Model simulations suggest that this reliance on separable filters is optimal given the limited spatial resolution of auditory input.
摘要:
运动知觉是一项基本的感官任务,起着至关重要的进化作用。在视觉上,传统上使用运动能量模型来描述运动处理,该模型具有时空不可分离的滤波器,该滤波器适用于捕获由移动对象提供的空间位置随时间的平滑连续变化。然而,目前还不清楚的是,作为听觉运动辨别基础的滤波器是否也是连续的运动检测器,还是通过比较随时间的离散声音位置(时空可分离)来推断运动。我们使用了心理物理逆相关范式,参与者在存在时空噪声的情况下辨别运动信号的方向,以确定作为听觉运动辨别基础的滤波器是时空可分离的还是不可分离的。然后,我们检查了这些听觉运动滤波器是否由于早期失明而改变。我们发现,有视力和早期失明的人都有可分离的过滤器。然而,早期失明的人对听觉运动的敏感性增加,降低对噪声的敏感性和在检测运动开始/偏移时更准确的滤波器。模型模拟表明,在听觉输入的空间分辨率有限的情况下,这种对可分离滤波器的依赖是最佳的。