Suppr超能文献

对视觉对象的尺寸和类别变化的识别。

Recognition of changes in the dimensions and categories of visual objects.

作者信息

Stefanova N

出版信息

Acta Physiol Pharmacol Bulg. 1979;5(1):27-33.

PMID:495132
Abstract

The aim of the present work is a comparison between recognition with changes in the dimensions of the objects and recognition with changes in the category of the objects. After preliminary training under tachistoscopic ocnditions, to a control set of different contour drawings are added: (1) the same drawings increased or desreased several times; (2) different new drawings with the same size. The percentage of recognized drawings is determined for such exposure time which is needed for the recognition of 60--80 per cent of the drawings in the control set. Recognition is found to be deteriorated in the case of changes both in the dimensions and in the category of the objects, the deterioration being double for the objects with changed category. This fact serves as a basis for rejecting the hypothesis that considerable changes in the dimensions could create new objects for the visual system. It was also found that recognition of pre-trained drawings is either not influenced or comparatively least influenced by the changes in the dimensions and categories. This result is probably due to the specificity of recognition of long trained objects. The results obtained do not contradict the assumption of the participation of spatial consecutive scanning of the iconic memory in the recognition of the spatial properties of visual objects.

摘要

本研究的目的是比较对物体尺寸变化的识别和对物体类别变化的识别。在速示条件下进行初步训练后,向一组不同的轮廓图中加入:(1)相同的图按比例放大或缩小若干倍;(2)相同尺寸的不同新图。针对识别对照组中60%至80%的图所需的曝光时间,确定识别图的百分比。结果发现,无论是物体尺寸还是类别发生变化,识别都会变差,且物体类别变化时变差程度加倍。这一事实成为拒绝尺寸大幅变化会为视觉系统创造新物体这一假设的依据。还发现,预先训练的图的识别要么不受尺寸和类别变化的影响,要么受到的影响相对最小。这一结果可能归因于对长期训练物体识别的特殊性。所得结果与关于图像记忆的空间连续扫描参与视觉物体空间属性识别的假设并不矛盾。

文献检索

告别复杂PubMed语法,用中文像聊天一样搜索,搜遍4000万医学文献。AI智能推荐,让科研检索更轻松。

立即免费搜索

文件翻译

保留排版,准确专业,支持PDF/Word/PPT等文件格式,支持 12+语言互译。

免费翻译文档

深度研究

AI帮你快速写综述,25分钟生成高质量综述,智能提取关键信息,辅助科研写作。

立即免费体验