A stretchable pressure sensor can provide highly sensitive and strain-unperturbed measurement to pressures on deformed skins.

Flexible pressure sensing plays a critical role in the human–machine interaction; therefore highly sensitive pressure sensors have promising potential in such applications. Capacitive pressure sensors have inherent attributes of simple configuration and fast response time. However, the existing applications of sensors are limited by the limited sensitivity as well as complex manufacturing. Herein, a highly sensitive ionic pressure sensor with the microstructured dielectric layer has been proposed. The polycarbonate membrane is used as the template to manufacture the microstructure of the concave meniscus. By packing the top electrode, dielectric layer and bottom electrode, the ionic pressure sensor is fabricated. The sensor presents a sensitivity of 35.96 k Pa ⁻¹ with an excellent robustness of tolerating repeated pressure for at least 200 000 cycles without fatigue. In addition, the response time of the sensor is 21 ms and the limit of detection is 0.61 Pa. The ionic pressure sensor offers a highly sensitive sensing device for robotics and human–machine interaction. Furthermore, the concave meniscus realized by the template-based method provides a novel guideline for other microstructure-enabled devices.

Graph convolutional networks (GCNs) have achieved outstanding performances on skeletonbased action recognition. However, several problems remain in existing GCN-based methods, and the spatialtemporal features are not discriminative enough. Temporal convolution with one fixed kernel cannot obtain more discriminative temporal features for different actions. Besides, only a single-scale feature is used for classification, which ignores the multilevel information. In this paper, we propose a novel multi-scale and multi-stream improved graph convolutional network (MM-IGCN). In each spatial-temporal block of MMIGCN, we employ an improved temporal convolution with multiple parallel kernels to enhance the temporal features. An improved GCN and an enhanced attention module are adopted in the block to strengthen spatialtemporal features. A multi-scale structure is first introduced in action recognition to obtain the multilevel information. The improved spatial-temporal blocks and multi-scale structure compose our single-stream model. Moreover, we adopt the bone cosine distance as a novel input feature. Five streams (joint, bone, their motions, and bone cosine distance) of features are fed into our single-stream model respectively, which compose our MM-IGCN. Experiments on two large datasets, NTU-RGB+D and NTU-RGB+D-120, illustrate that our single-stream model achieves state-of-the-art, and our MM-IGCN is far superior to other models.