Collaborative Research: Seeing Science: Using Computer Vision to Explore the Scientific Principles Behind Everyday Objects

理解科学对于培养学生理解周围的世界、做出明智的决定、参与公民社会和劳动力至关重要。然而，对许多年轻人来说，科学是一个神秘的知识体系，感觉与他们的生活脱节。该项目旨在将科学带入中学生的家中，让他们看到日常物品背后的科学，并将生活环境转变为引人入胜的学习空间。学生们将在手机上进行基于探究的学习单元，探索诸如扩散、电力和厨房、卧室和当地公园中存在的简单机器等STEM现象。他们将能够拍摄他们的家和邻居的照片和视频，计算机视觉算法将用图表、模型和模拟来增强这些图像，这些图表、模型和模拟说明了解释STEM现象的原理和机制。这些叠加将允许学生观察、实验和预测诸如茶叶在热水中扩散或热量穿过墙壁等现象。该项目将利用现有的技术设备，如照相手机，创造“镜头”，使学生能够看到他们周围的科学。通过致力于这种低成本的解决方案，该项目旨在使更多的学生能够接受科学教育，并使在家学习成为可能，同时避免对家庭资源造成不必要的压力。在多样化、低资源的环境中运行，激发了计算机视觉的基本进步：首先，算法必须自动建立给定物理现象的场景的3D和时间表示。其次，系统必须为教育工作者提供挂钩，以决定哪些图形应该在哪个时间和位置覆盖在这个场景上。此外，该项目将致力于以人为本的设计，以易于获取、易于使用的方式将尖端技术带给年轻人，并实现教育目标。调查人员将对家长、学生和老师进行广泛的访谈，了解他们愿意与研究人员、同龄人和老师分享的学生校外生活的各个方面。这些数据将使该团队认识到建立在学生生活和文化基础上的公平科学教育的好处。这项研究将有助于促进发展一种更主动、更包容的在家参与科学探究的方式，鼓励学生从小就与科学建立个人联系。对于那些认为科学与他们的生活脱节的风险最大的学生来说，这一点尤其重要，他们可以从看到科学在他们的生活和社区中发挥作用中获益最多。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Understanding science is critical for preparing students to make sense of the world around them, make informed decisions, and participate in civic society and in the workforce. However, for many youth, science is a mysterious body of knowledge that feels disconnected from their lives. This project aims to bring science into middle school students’ homes, allowing them to see the science behind everyday objects and transforming lived environments into engaging learning spaces. Students will work on inquiry-based learning units on mobile phones that explore STEM phenomena topics like diffusion, electricity, and simple machines that are present in their kitchens, bedrooms, and local parks. They will be able to take photos and videos of their home and neighborhood, and computer vision algorithms will augment these images with diagrams, models, and simulations that illustrate the principles and mechanisms that explain the STEM phenomena. These overlays will allow students to observe, experiment with, and make predictions for phenomena such as tea diffusing in hot water or heat traveling through walls. The project will capitalize on existing technological devices, such as camera phones, to create “lenses” which enable students to see the science that is all around them. By committing to such low-cost solutions, the project aims to make science education accessible to more students, and enable at-home learning while avoiding undue pressures on family resources. Operating in diverse, low-resource environments motivates fundamental advances in computer vision: First, algorithms must automatically build up a 3D and temporal representation of a scene of a given physical phenomenon. Second, the system must expose hooks for educators to decide which graphics should be overlaid at which time and in which place atop this scene. Further, the project will engage in human-centered design to bring cutting-edge technologies to youth in ways that are accessible, easy to use, and achieve educational goals. Investigators will conduct extensive interviews with parents, students, and teachers about the aspects of students’ out-of-school lives that they would be willing to share with researchers, peers, and teachers. These data will enable the team to realize the benefits of equitable science education that builds on students’ lives and cultures. This research will help foster the development of a more agentic, inclusive way of engaging in science inquiry at home, encouraging students to have a personal connection with science from a young age. This is particularly important for students most at risk to perceive science as disconnected from their lives, and whom can benefit most from seeing science at work in their lives and community.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.