No abstract

Piezoelectric micromachined ultrasonic transducers (PMUTs) are used to receive and transmit ultrasonic signals in industrial and biomedical applications. This type of transducer can be miniaturized and integrated with electronic systems since each element is small and the power requirements are low. The bandwidth of the PMUT may be narrow in some conventional designs; however, it is possible to apply modified structures to enhance this. This paper presents a methodology for improving the bandwidth of air-coupled PMUTs without sensitivity loss by connecting a number of resonating pipes of various lengths to a cavity. A prototype piezoelectric diaphragm ultrasonic transducer is presented to prove the theory. This novel device was fabricated by additive manufacturing (3-D printing), and consists of a polyvinylidene fluoride thin film over a stereolithography designed backplate. The backplate design is inspired by a pipe organ musical instrument, where the resonant frequency (pitch) of each pipe is mainly determined by its length. The -6-dB bandwidth of the "pipe organ" air-coupled transducer is 55.7% and 58.5% in transmitting and receiving modes, respectively, which is ~5 times wider than a custom-built standard device.

In this paper, we build a model named VuLASTE, which regards vulnerability detection as a special text classification task. To solve the vocabulary explosion problem, VuLASTE uses a byte level BPE algorithm from natural language processing. In VuLASTE, a new AST path embedding is added to represent source code nesting information. We also use a combination of global and dilated window attention from Longformer to extract long sequence semantic from source code. To solve the data imbalance problem, which is a common problem in vulnerability detection datasets, focal loss is used as loss function to make model focus on poorly classified cases during training. To test our model performance on real-world source code, we build a cross-language and multi-repository vulnerability dataset from Github Security Advisory Database. On this dataset, VuLASTE achieved top 50, top 100, top 200, top 500 hits of 29, 51, 86, 228, which are higher than state-of-art researches.

No abstract

No abstract