Gesture Recognition Software

About

Track Code: TAMU 4436 Tags   Smart Devices, context aware, gesture, recognition, Dynamic Time Warping, Contextual awareness, awareness, control, wearable, wearables, devices, sensor Overview Home and office devices are becoming increasingly connected through being equipped with sensors and wireless connectivity. This “Internet of Things”-like connectivity allows for more possibilities for intuitive interaction with users. The invention described here can be added to any such device to seamlessly enable user interactions: proposed is a solution and algorithms that can be implemented on low cost wearable devices to enable greater user-device interaction functionality. Technology By leveraging wrist-worn wearable technology, this system uses a multi-Hidden Markov Model (HMM) to recognize gestures robustly and independently of the sensor’s orientation on the wrist. Thus, in addition to simply recognizing a gesture, the system uses the wearable device and environmental sensors to enable “contextual awareness” and significantly improve gesture recognition.                    Advantages Greatly improved accuracy when compared to current methods such as Dynamic Time Warping (DTW) – shown in the table above Determines which device the user intends to control through contextual awareness Replaces control devices, appliances, and applications with a single, wearable device Applications Gesture commands to control smart devices Stage of Development Developed and tested rigorously with comparative data against other technologies Patent Status Patent Pending Publications K. Liu, C. Chen, R. Jafari, N. Kehtarnavaz, “Multi-HMM classification for hand gesture recognition using two differing modality sensors.” The 10th IEEE Dallas Circuits and Systems Conference (DCAS), Oct. 12-13, 2014. K. Liu, C. Chen, R. Jafari, N. Kehtarnavaz, “Fusion of Inertial and Depth Sensor Data for Robust Hand Gesture Recognition.” IEEE Sensors Journal (SJ), v. 14, 1898-1903. June, 2014.