Abstract
Gesture recognition is the most intuitive form of human computer-interface. Gesture sensing can replace interfaces such as touch and clicks needed for interacting with a device. Gesture recognition in a practical scenario is an open-set classification, i.e. the recognition system should classify correct known gestures while rejecting arbitrary unknown gestures during inference. To address the issue of gesture recognition in an open set, we present, in this paper, a novel distance-metric based meta-learning approach to learn embedding features from a video of range-Doppler images generated by hand gestures at the radar receiver. Further, k-Nearest Neighbor (kNN) is used to classify known gestures, distance-thresholding is used to reject unknown gesture motions and clustering is used to add new custom gestures on-the-fly without explicit model re-training. We propose to use 3D Deep Convolutional Neural Network (3D-DCNN) architecture to learn the embedding model using distance-based triplet-loss similarity metric. We demonstrate our approach to correctly classify gestures using short-range 60-GHz compact short-range radar sensor achieving an overall classification accuracy of 94.5% over six fine-grained gestures under challenging practical environments, while rejecting other unknown gestures with 0.935 F1 score, and capable of adding new gestures on-the-fly without an explicit model re-training.
Highlights
Dynamic gestures are one of the most intuitive and effective approach for human-computer interaction
Hand gesture recognition systems have been based on optical sensors and cameras [3]
Triplet selection or triplet mining is a critical aspect of training a deep convolutional neural network (DCNN) using triplet loss, if not done correctly the loss can get stuck in local minima after reducing drastically in the first few epochs
Summary
Dynamic gestures are one of the most intuitive and effective approach for human-computer interaction. In [4], the authors propose a novel 3D deep convolutional neural network (3D-DCNN) for feature extraction, long-short term memory (LSTM) with connectionist temporal classification (CTC) loss function for the recognition of a gesture across time on camera and depth data. The embedding model, can simultaneously minimize the distance between similar gestures and maximize the distance between different gestures with triplet loss function metric Such model and metrics have been successfully applied to a 2D image in literature, to the best of author’s knowledge this is first work that extends it to 3D data, video of radar RDIs. With the well-learned gesture features embedding, k-Nearest Neighbor (kNN) algorithm is used to recognize a known gesture even under an alien environment while rejecting unknown gestures using a simple thresholding technique to minimize false alarms.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
