Abstract
AbstractThe Gestalt laws of perceptual organization, which describe how visual elements in an image are grouped and interpreted, have traditionally been thought of as innate. Given past research showing that these laws have ecological validity, we investigate whether deep learning methods infer Gestalt laws from the statistics of natural scenes. We examine the law of closure, which asserts that human visual perception tends to “close the gap” by assembling elements that can jointly be interpreted as a complete figure or object. We demonstrate that a state-of-the-art convolutional neural network, trained to classify natural images, exhibits closure on synthetic displays of edge fragments, as assessed by similarity of internal representations. This finding provides further support for the hypothesis that the human perceptual system is even more elegant than the Gestaltists imagined: a single law—adaptation to the statistical structure of the environment—might suffice as fundamental.
Publisher
Springer Science and Business Media LLC
Subject
Developmental and Educational Psychology,Neuropsychology and Physiological Psychology
Reference65 articles.
1. Amanatiadis, A., Kaburlasos, V.G., & Kosmatopoulos, E.B. (2018). Understanding deep convolutional networks through Gestalt theory. In IEEE International conference on imaging systems and techniques (IST) (pp. 1–6). Krakow: IEEE Press.
2. Baker, N., Kellman, P.J., Erlikhman, G., & Lu, H. (2018). Deep convolutional networks do not perceive illusory contours. In Proceedings of the 40th Annual conference of the cognitive science society, cognitive science society, Austin, TX (pp. 1310–1315).
3. Bau, D., Zhou, B., Khosla, A., Oliva, A., & Torralba, A. (2017). Network dissection: Quantifying interpretability of deep visual representations. In Computer vision and pattern recognition.
4. Behrens, R.R. (1998). Art, design and Gestalt theory. Leonardo, 31(4), 299–303.
5. Bender, L. (1938). A visual motor Gestalt test and its clinical use. Research Monographs, American Orthopsychiatric Association.
Cited by
19 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献