Fully Flexible Smart Gloves and Deep Learning Motion Intention Prediction for Ultralow Latency VR Interactions

Most reported flexible smart gloves are complex in mass processing and have stability problems in the interface with the circuit, which seriously limits their wide application. Moreover, the communication latency caused by wireless transmission is also a factor that seriously restricts remote interaction and simply improving the signal transmission speed has a bottleneck. Here, an integrated full print production flexible smart glove and an advanced response method based on deep learning motion intention prediction were developed to overcome these shortcomings. All device components are integrated on a flexible printed circuit board, including a topological carbon-silver strain sensor, a serpentine stretchable wire, and a wireless signal circuit board, which is suitable for mass production and has specific stability and stretchability. An optimized model based on long short-term memory is designed to predict finger motion intention and respond 100-600 ms in advance to reduce communication latency. This letter proposes a flexible smart glove that is suitable for mass production and provides a new way to solve the remote interaction latency.

Automated summary

Most reported flexible smart gloves are complex in mass processing and have stability problems in the interface with the circuit, limiting their wide application. The communication latency caused by wireless transmission also restricts remote interaction, and improving signal transmission speed has limitations. In this letter, an integrated full print production flexible smart glove and a deep learning based motion intention prediction method were developed to overcome these shortcomings and reduce communication latency.