CompCog: Bridging the gap between behavioral and neural correlates of attention using a computational model of neural mechanisms

我们的视觉系统面临的一个挑战是能够专注于对当前任务重要的信息，同时还能对意外事件做出反应。例如，开车时，如果你仔细寻找街道标志，你也应该能够停下来，以避免汽车从小路上驶出。为了应对这一挑战，我们的大脑使用了一个注意力系统，该系统监控环境中的重要信息，并决定在每个时刻应该关注什么。这项拟议的研究将使用检测行为和大脑活动的实验，以及整合了许多不同实验室的其他发现的计算机模型来研究这种注意力系统。这个模型将帮助研究人员更好地理解脑电波是如何与脑细胞的活动和视觉注意时的行为相关的。该模型将以任何人都可以在自己的电脑上运行的格式公开提供，这样其他研究人员就可以使用它来帮助我们在许多重要领域加深理解，比如汽车和其他自动化系统将如何能够安全地导航并与世界互动。最后，这项工作支持旨在让本科生以高度技术性的方法进行科学研究的教育努力，并资助高中的外展努力，让学生有机会培养对神经科学和机器人的更大兴趣。这项工作建立在对视觉注意力性质的数十年研究的基础上，这些研究产生了大量关于大脑如何处理信息的数据。在这项工作的基础上，研究人员创建了一个注意力计算模型，试图模拟大脑如何选择要参与的信息片段。为了验证这一模型，他们将进行一系列八项新的实验，以测试该模型的具体预测。这些测试将提供关于模型中不同层的神经元应该如何通信的信息，以最接近人类大脑的注意力系统。最终的模型将被汇编成一个版本，可以由其他研究人员或教育工作者下载。这个模型将提供一个精致的图形用户界面，允许新手用户探索模拟的注意力系统是如何工作的，以及脑电波是如何产生的。另一个目标是开发一种新的实验，测试视觉和注意力之间的延迟。来自这一范例的数据将为科学家提供至关重要的细节，了解人类视觉系统如何暂时保存信息，同时确定如何处理这些信息。

A challenge for our visual system is being able to focus on information important for our current task, while also being responsive to unexpected events. For example, when driving, if one is looking carefully for a street sign, one should also be able to stop to avoid a car pulling out from a side street. To address this challenge, our brain uses an attentional system that monitors the environment for important information and decides what should be focused at each moment in time. The proposed research will investigate this attentional system, using experiments that examine behaviors and brain activity, as well as with a computer model that integrates other findings from many different labs. This model will help researchers to better understand how brainwaves are related to both the activity of brain cells and behavior when visual attention is engaged. The model will be made publicly available in a format that anyone can run on their own computer, so that other researchers can use it to help advance our understanding in many important areas, such as how cars and other automated systems will be able to navigate and interact safely with the world. Finally, this work supports educational efforts aimed at engaging undergraduate students with a highly technical approach to scientific investigation and funds outreach efforts in high schools to give students the opportunity to develop a greater interest in neuroscience and robotics.This work builds on decades of research into the nature of visual attention, which has generated a large volume of data about how the brain processes information. Based on this work, the researchers have created a computational model of attention that attempts to simulate how the brain chooses which pieces of information to attend. To validate this model, they will conduct a series of eight new experiments that test specific predictions of the model. These tests will provide information about how the different layers of neurons in the model should communicate to most closely approximate the human brain's attentional system. The final model will be compiled into a version that can be downloaded by other researchers or educators. This model will provide a polished graphical user interface, allowing novice users to explore how the simulated attention system works, and how brainwaves are generated. A further objective will be to develop a new kind of experiment that tests the delay between vision and attention. The data from this paradigm will give scientists crucial details about how the human visual system temporarily holds information while determining what to do with it.