STTR Phase II: Dynamic Robust Hand Model for Gesture Intent Recognition

该项目更广泛的影响/商业潜力来自于解决VR/AR(预计到2020年增长到1,500亿美元)、机器人和物联网(到2019年分别为1,350亿美元和到2020年1，7T美元)等基于手势的重要输入挑战。如果这项技术成功地降低了高技术风险，代表着手势识别3D手模型技术水平的巨大飞跃，并有可能成为AR、VR、机器人和物联网应用的行业标准，在教育、医疗和医疗保健领域具有广泛的社会影响。通过更好地沟通、学习和适应他们的互动需求，它在服务于残疾人社区提高他们的生活质量方面的潜力进一步放大了其更广泛的影响。这个小型企业技术转移(STTR)第二阶段项目旨在通过开发用于手势意图识别的动态手势跟踪模型来显著促进当前的3D手势识别技术。它对遮挡具有很强的健壮性，并且对相机方向和位置的变化具有容忍性。这项研究将带来一次革命性的飞跃，超越学术和商业手模型的现状，并克服迄今被证明难以克服的关键技术障碍。它解决了以下关键挑战，并涉及非常高的技术风险：1)在持有目标对象的同时进行健壮的手部跟踪；2)在不使用任何基准标记的情况下使用对象进行健壮的有形交互2)可穿戴的低调手部用于触摸交互检测。这个第二阶段的项目将通过1)数据采集和手部物体姿态估计，2)通过增强的有形交互来理解用户意图，以及3)系统验证和用户测试来实现这些目标。

The broader impact/commercial potential of this project stems from addressing theimportant hand gesture based input challenges of VR/AR (expected to grow to $150B by2020), Robotics and IoT ($135B by 2019 and $1.7T by 2020 respectively). This technology,if successful in mitigating the high technical risks, represents a huge leap in the state of theart in 3D hand models for gesture recognition and has the potential to be the industrystandard for AR, VR, Robotics and IoT applications with broad societal impact in education,medical and healthcare. Its broader impact is further amplified by the potential in serving theneeds of the disabled community in improving their quality of life by being better able tocommunicate, learn and adapt to their interaction needs.This Small Business Technology Transfer (STTR) Phase 2 project aims to significantlyadvance current 3D hand gesture recognition technology by developing a dynamic handtracking model for gesture intent recognition. It is robust against occlusion and tolerant tovariations in camera orientation and position. This research will result in a transformativeleap above the current state of academic and commercial hand models and overcome keytechnical hurdles that have so far proven difficult to overcome. It solves the following keychallenges and involves very high technical risks: 1) robust hand tracking while holdingobjects and 2) robust tangible interactions using objects without using any fiducial markers2) low profile hand wearable for touch interaction detection. This Phase 2 project willachieve these objectives by 1) data acquisition and hand-object pose estimation, 2)understanding user intents with enhanced tangible interactions, and 3) system validationand user testing.