Learning diagnostic latent representations for human material perception: common mechanisms and individual variability

学习人类物质感知的诊断潜在表征：共同机制和个体差异

基本信息

批准号：
10580295
负责人：
Bei Xiao
金额：
$ 42.05万
依托单位：
AMERICAN UNIVERSITY
依托单位国家：
美国
项目类别：
财政年份：
2023
资助国家：
美国
起止时间：
2023-04-01 至 2026-03-31
项目状态：
未结题

来源：
https://reporter.nih.gov/project-details/10580295
关键词：
3-Dimensional 3D Print Affect American Appearance Artificial Intelligence Biological Categories Classification Cognition Collaborations Color Computers Cues Data Data Set Dental Diagnostic Diagnostic Imaging Dimensions Discrimination Emerging Technologies Environment Esthetics Evolution Food Food Selections Fruit Geometry Glass Goals Human Human Subject Research Image Individual Individual Differences Judgment Label Learning Lighting Machine Learning Measures Meat Mentorship Methods Modeling Neurosciences Ocular Prosthesis Operative Surgical Procedures Outcome Paper Perception Persons Process Property Psychophysics Publishing Quality of life Research Role Semantics Shapes Skin Structure Students Surface Surface Properties Testing Texture Training Universities Variant Vision Visual Visual System Visual impairment Walking Waxes Work computer science deep learning deep neural network educational atmosphere experience falls generative adversarial network graduate student grasp human subject image processing individual variation inter-individual variation multidisciplinary multisensory neural correlate novel novel strategies object recognition peer physical property prevent programs public health relevance recruit statistical learning statistics tool training opportunity undergraduate student unsupervised learning visual processing

项目摘要

Abstract Visually discriminating and identifying materials (such as judging whether a cup is made of plastic or glass) is crucial for everyday tasks, such as walking on different surfaces, using tools, and selecting food; and yet material perception remains poorly understood. The main challenge is that a given material can take an enormous variety of appearances depending on the 3D shape, lighting, and object class, and humans must untangle these to achieve perceptual constancy. Previous research revealed useful image cues and found that 3D geometry interacts with the material perception in intricate ways. The discovered image cues, however, do not generalize across materials and scenes. The proposed work will combine unsupervised generative models with human psychophysics to identify a representation that can disentangle physical properties and discover diagnostic image features without labeled image data. The speciﬁc Aim 1 is to identify a latent representation that predicts human material discrimination, using unsupervised deep neural networks trained with computer rendered images. The speciﬁc Aim 2 is to characterize high-level semantic material perception, the effects of high-level recognition as well as individual differences on attribute rating and recognition tasks. To discover a representation of real-world materials, the PI and the team will train a unsupervised style-based Generative Adversarial Network (StyleGAN) on real-world photographs. The preliminary results show that StyleGAN can generate realistic and diverse images of materials. Collectively, these studies will explore how the semantic-level material perception process relates to the statistical structure of the natural environment learned from unsupervised models. The proposed work will also uncover the task-dependent interplay between high-level vision and mid-level representations, and provide guidance for seeking neural correlates of material perception. The methods developed in this proposal, such as discovering perceptual dimensions with limited human labeled data and characterizing individual variability, have impact for other research in cognition. The AREA proposal provides a unique multidisciplinary training opportunity to engage diverse undergraduate students at American University in the research of psychophysics, machine learning, and image processing. The PI and students will also investigate a novel method of recruiting under-represented human subjects using "peer-recruiting." Finally, the expected ﬁndings of this proposal will have implications for the long-standing debate about the degree to which perceptual representations are predetermined by evolution or learned via experience.

抽象的视觉辨别和识别材质（例如判断杯子是塑料的还是玻璃的）对于日常任务至关重要，例如在不同的表面上行走、使用工具和选择食物；物质感知仍然知之甚少，主要的挑战是给定的材料可能需要巨大的影响。各种外观取决于 3D 形状、照明和对象类别，人类必须理清这些先前的研究揭示了有用的图像线索并发现 3D 几何形状。然而，所发现的图像线索并不能一概而论。拟议的工作将把无监督生成模型与人类结合起来。心理物理学来识别可以解开物理特性并发现诊断图像的表示没有标记图像数据的特征具体目标 1 是识别预测人类的潜在表示。材料歧视，使用由计算机监督渲染图像训练的无监督深度神经网络。具体目标 2 是表征高级语义材料感知，高级识别的效果为以及属性评级和识别任务上的个体差异，以发现现实世界的表示。材料、PI 和团队将训练一个无监督的基于风格的生成对抗网络（StyleGAN）初步结果表明 StyleGAN 可以生成逼真且多样化的图像。总的来说，这些研究将探讨语义层面的材料感知过程如何相互关联。所提出的工作将是从无监督模型中学习到的自然环境的统计结构。还揭示了高层视觉和中层表征之间依赖于任务的相互作用，并提供本提案中开发的方法，例如寻找物质感知的神经关联的指导。用有限的人类标记数据发现感知维度并表征个体差异，对其他认知研究产生影响 AREA 提案提供了独特的多学科培训。让美国大学的不同本科生参与心理物理学研究的机会，机器学习和图像处理 PI 和学生还将研究一种新的招募方法。最后，该提案的预期结果将是：对关于感知表征在多大程度上预先确定的长期争论的影响通过进化或通过经验学习。