Abstract:The uncertainty of hand gestures, the variability of gestures across subjects, and the high cost of collecting a large amount of annotated data lead to a great challenge to the robust recognition of gestures, and thus it remains quite crucial to capture the informative features of hand movements and to mitigate inter-subject variations. To this end, we propose a gesture recognition model that uses two different types of sensors and optimizes the feature space towards enhanced accuracy and better generalization. Specifically, we use an accelerometer and a surface electromyography sensor to capture kinematic and physiological signals of hand movements. We use a sliding window to divide the streaming sensor data and then extract time-domain and frequency-domain features from each segment to return feature vectors. Afterwards, the feature space is optimized with a feature selector and a gesture recognizer is optimized. To handle the case where no labeled training data are available for a new user, we apply the transfer learning technique to reuse the cross-subject knowledge. Finally, extensive comparative experiments concerning different classification models, different sensors, and different types of features are conducted. Results show that the joint use of kinematic and physiological sensors generally outperforms the use of single sensor, indicating the synthetic effect of different sensors, and that the use of transfer learning helps improve the cross-subject recognition accuracy. In addition, we quantitatively investigate the impact of null gesture on a gesture recognizer and results indicate that null gesture would lower its accuracy, enlightening related studies to consider it.

Accelerometer Based Gesture Recognition Using Fusion Features and SVM.

Gesture Recognition Based on 3D Accelerometer for Cell Phones Interaction

Gesture Recognition with a 3-D Accelerometer

A New Feature Fusion Method for Handwritten Character Recognition Based on 3d Accelerometer

Accelerometer-based Gesture Recognition Using MFCC and HMM

Gesture Recognition Using Inertial Sensors with 1D Convolutional Neural Network

Feature Fusion of sEMG and Ultrasound Signals in Hand Gesture Recognition

Motion-Sensor Fusion-Based Gesture Recognition and Its Vlsi Architecture Design for Mobile Devices

A Hybrid Multimodal Fusion Framework for Semg-Acc-based Hand Gesture Recognition

Fusion of kinematic and physiological sensors for hand gesture recognition

Dynamic Gesture Recognition Based on Multiple Sensors Fusion Technology

Hand Gesture Recognition using Deep Feature Fusion Network based on Wearable Sensors

WristCam: A Wearable Sensor for Hand Trajectory Gesture Recognition and Intelligent Human–Robot Interaction

Multi-Hand Gesture Recognition Using Automotive FMCW Radar Sensor

Dynamic hand gesture recognition using hidden Markov models

Millimeter wave gesture recognition using multi-feature fusion models in complex scenes

MGRA: Motion Gesture Recognition Via Accelerometer

Finger Gesture Recognition Based on 3D-Accelerometer and 3D-Gyroscope

A Hand Gesture Recognition Framework and Wearable Gesture-Based Interaction Prototype for Mobile Devices.

Sensor fusion using EMG and vision for hand gesture classification in mobile applications

Dynamic Gesture Recognition Based on Feature Fusion Network and Variant ConvLSTM.