Suppr超能文献

大脑中如何计算和表示分割和绑定?

How are segmentation and binding computed and represented in the brain?

机构信息

Frankfurt Institute for Advanced Studies, Ruth-Moufang-Str. 1, 60476, Frankfurt, Germany.

Institute for Neural Computation, UZH/ETH, Winterthurerstrasse 190, 8057, Zurich, Switzerland.

出版信息

Cogn Process. 2024 Aug;25(Suppl 1):67-72. doi: 10.1007/s10339-024-01214-z. Epub 2024 Aug 10.

Abstract

A programmatically described solution to the segmentation problem is taken as opportunity to dicuss the neural architecture problem of vision. At the center of this problem is the formation of holistic entities (the Gestalt phenomenon) out of masses of neurons (the binding problem). As formulated in the Dynamic Net Architecture (DNA), neurons can become part of a (short-term) stable state only if supported inside a coherent network ('net'). Integration into nets is the basis for global treatment of structures, such as recognition as rigid body or projection to an invariant model.

摘要

我们将一个程序化描述的分割问题解决方案作为讨论视觉神经架构问题的契机。这个问题的核心是如何从大量神经元中形成整体实体(格式塔现象)(即绑定问题)。正如动态网络架构(DNA)所表述的,只有神经元在一个连贯的网络(“网络”)中得到支持,才能成为(短期)稳定状态的一部分。神经元整合到网络中是对结构进行全局处理的基础,例如识别为刚体或投影到不变模型。

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验