Fusion network for small target detection based on YOLO and attention mechanism

Target detection is an important task in computer vision research, and such an anomaly detection and the topic of small target detection task is more concerned. However, there are still some problems in this kind of researches, such as small target detection in complex environments is susceptible to background interference and poor detection results. To solve these issues, this study proposes a method which introduces the attention mechanism into the you only look once(YOLO) network. In addition, the amateur-produced mask dataset was created and experiments were conducted. The results showed that the detection effect of the proposed mothed is much better.

Mechanism of CeMgAl_(11)O_(19):Tb^(3+) alkaline fusion with sodium hydroxide

Comprehensive CeMgA111O19： Tb3＋ （CTMA） disintegration via alkaline fusion was discussed. The rare earth （RE） elements in CTMA were dissolved by HC1 completely after alkaline fusion. Relationships between the alkaline fusion temperature and various properties of the compounds were examined by various techniques to elu- cidate their roles in the expected CTMA disintegration. X-ray diffraction （XRD） analysis indicates the phase transformation sequence. A scientific hypothesis of crystal structure disintegration presents that sodium ions substitute for the europium and barium ions in the mirror plane and magnesium ions in the spinel block successively, resulting in that more oxygen vacancies and interstitial sodium ions appear. The unit cell [P63/mmc （194）] breaks from the mirror plane. Then it is decomposed into NaA102, and magnesium, cerium, and terbium ions combine with free OH- into MgO, Tb2O3 and CeO2; Tb2O3 and CeO2 change into Ceo.6Tbo.O2-x. In the end, the rare earth oxide is recycled easily by the acidolysis. The mechanism provides fundamental basis for recycling of REEs from waste phosphors.

A life-prediction method for lithium-ion batteries based on a fusion model and an attention mechanism

The current life-prediction models for lithium-ion batteries have several problems, such as the construction of complex feature structures, a high number of feature dimensions, and inaccurate prediction results. To overcome these problems, this paper proposes a deep-learning model combining an autoencoder network and a long short-term memory network. First, this model applies the characteristics of the autoencoder to reduce the dimensionality of the high-dimensional features extracted from the battery data set and realize the fusion of complex time-domain features, which overcomes the problems of redundant model information and low computational efficiency. This model then uses a long short-term memory network that is sensitive to time-series data to solve the long-path dependence problem in the prediction of battery life. Lastly, the attention mechanism is used to give greater weight to features that have a greater impact on the target value, which enhances the learning effect of the model on the long input sequence. To verify the efficacy of the proposed model, this paper uses NASA's lithium-ion battery cycle life data set.

New construction for expert system based on innovative knowledge discovery technology

Knowledge acquisition Is the bottleneck of expert system. To solve this problem, KD （D＆K）, which is a comprehensive knowledge discovery process model coopersting both database and knowledge base, and related technology are proposed. Then based on KD （D＆K） and related technology, the new construction of Expert System based on Knowledge Discovery （ESKD） Is proposed. As the key knowledge acqulsltlon component of ESKD, KD （D＆K） Is composed of KDD＊ and KDK＊. KDD＊- the new process model based on double bases cooperating mechanism; KDK＊- the new process model based on double-basis fusion mechanism are Introduced, respectively. The overall framework of ESKD Is proposed. Some sub-systems and dynamic knowledge base system are discussed. Flnelly, the effectiveness and advantages of ESKD are tested In a real-world agriculture database. We hope that ESKD may be useful for the new generation of expert systems.