Acquisition of an Integrated Computational and Psychophysical Laboratory

最近成立的麻省理工学院生物和计算学习中心(CBCL)联合了七名计算神经科学研究人员，他们的专业领域包括视觉感知、运动控制、语言和学习理论。最近，提供了空间，并对其进行了翻新，以将该小组聚集在一个单一的物理位置，该NSF奖通过购买一个现代化的联网计算机实验室，包括工作站、服务器和网络中心，为研究和培训活动提供了基础设施。将购买的物品包括3 Bay Networks Network Huns、3台Sun UltraServer、一台SGI Indigo High-Impact Workstation和一台Electrhome Marquee 8500投影仪。该设施的核心是一个虚拟现实系统，在这个系统中，复杂的多模式刺激(视觉、听觉和触觉)可以被合成并呈现给受试者。一个记录受试者四肢和眼睛位置的系统将被开发出来，与刺激显示系统紧密结合。电子家选取景投影仪是一款高亮度投影仪，可以在大视场中显示刺激。视觉刺激将在高性能Silicon Graphics Indigo上生成，当配备High Impact视频板时，可以生成实时纹理映射图像。就智力在头脑和大脑中的算法实现而言，理解智力是现代科学面临的最大挑战之一。它有很多方面；其中一个已经成为整个计算和脑科学的统一目标的问题是理解动物或机器如何学习的问题。CBCL通过发展对学习的理论理解，并将这种理解应用于视觉感知、运动控制和语言领域的各种问题，在学习机制研究方面发挥了领导作用。通过这笔赠款提供的工具将大大推进这一目标。

The recently established MIT Center for Biological and Computational Learning (CBCL) unites seven faculty in the study of computational neuroscience with specialization in the areas of visual perception, motor control, language and the theory of learning. Recently space has been made available and renovated to bring the group together in a single physical location and this NSF award provides an infrastructure for research and training activities with the purchase of a modern networked computer laboratory, including workstations, servers and a network hub. Items to be purchased include 3 Bay Networks Network Huns, three Sun UltraServers, an SGI Indigo High-Impact Workstation and an Electrohome Marquee 8500 projector. Central to the facility is a 'virtual reality` system in which complex multi-modal stimuli (visual, auditory and haptic), can be synthesized and presented to subjects. A system for recording the positions of subjects' limbs and eyes will be developed, with tight coupling to the stimulus display system. The Electrohome Marquee projector is a high-brightness projector which displays stimuli in a large field of view. The visual stimuli will be generated on the high-performance Silicon Graphics Indigo which, when equipped with a High Impact video board, can generate real time texture-mapped images. The problem of understanding intelligence in terms of its algorithmic realization in the mind and brain is one of the greatest challenges of modern science. It has many facets; one which has become a unifying goal throughout the computational and brain sciences is the problem of understanding how an animal or a machine can learn. CBCL has provided leadership in research on learning mechanisms by developing a theoretical understanding of learning and by applying this understanding to a wide variety of problems in the domains of visual perception, motor control and language. The instrumentation provided through this grant will significantly further that goal.