In this work, we share our experience on tele-knowledge pre-training for fault analysis, a crucial task in telecommunication applications that requires a wide range of knowledge normally found in both machine log data and product documents. To organize this knowledge from experts uniformly, we propose to create a Tele-KG (tele-knowledge graph). Using this valuable data, we further propose a tele-domain language pre-training model TeleBERT and its knowledge-enhanced version, a tele-knowledge re-training model KTeleBERT. which includes effective prompt hints, adaptive numerical data encoding, and two knowledge injection paradigms. Concretely, our proposal includes two stages: first, pre-training TeleBERT on 20 million tele-related corpora, and then re-training it on 1 million causal and machine-related corpora to obtain KTeleBERT. Our evaluation on multiple tasks related to fault analysis in tele-applications, including root-cause analysis, event association prediction, and fault chain tracing, shows that pre-training a language model with tele-domain data is beneficial for downstream tasks. Moreover, the KTeleBERT re-training further improves the performance of task models, highlighting the effectiveness of incorporating diverse tele-knowledge into the model.

为提升烟草制丝加工、生产质量等数据的综合分析及应用能力,设计了一套面向制丝车间高数据吞吐的电能数据采集与分析系统.该系统基于协议网关和Cassandra数据仓库实现异构数据的汇聚交互;通过Spring Batch批处理技术及流式的、面向切面的任务管理实现任务创建、管理与异常分析;运用Quartz实现任务的定时调度;采用RabbitMQ消息中间件实现采集数据的消峰;通过MQTT协议与Vue前端技术将信息展示给用户,最终实现从数据调度、存储、分析到展示的全流程功能.测试结果表明,该系统访问延时小、承载能力和稳定性强,可以满足烟草企业制丝车间的电能数据分析需求.

Maritime ships detection is one of the main tasks in unmanned surface vehicle (USV)’s visual system. This paper proposes a kind of USV maritime ships detection al-gorithm based on object proposals. Firstly, a modified edge boxes algorithm is utilized to extract the edge information of the image, and an objectness score function is established to obtain object proposals. Secondly, a histogram of oriented gradient (HOG) feature model is built for the ship, and the support vector machine (SVM) is utilized to iteratively train a classifier by a bootstrap method. Finally, the feature descriptor of object proposals is fed into the classifier, and detecting the ship. In addition, the sea-sky line is utilized to further improve the detection performance of the algorithm based on the environment of USV. The experimental results show that the algorithm can rapidly and accurately detect the ship on the sea, and achieve a relatively high detection rate. And the algorithm has strong robustness to the change of the scale and the illumination conditions.

Integrating spectral and spatial information is proved effective in improving the accuracy of hyperspectral imagery classification. In recent studies, two kinds of approaches are widely investigated: (1) developing a multiple feature fusion (MFF) strategy; and (2) designing a powerful spectral-spatial feature extraction (FE) algorithm. In this paper, we combine the advantages of MFF and FE, and propose an ensemble based feature representation method for hyperspectral imagery classification, which aims at generating a hierarchical feature representation for the original hyperspectral data. The proposed method is composed of three cascaded layers: firstly, multiple features, including local, global and spectral, are extracted from the hyperspectral data. Next, a new hashing based feature representation method is proposed and conducted on the features obtained in the first layer. Finally, a simple but efficient extreme learning machine classifier is employed to get the classification results. To some extent, the proposed method is a combination of MFF and FE: instead of feature fusion or single feature extraction, we use an ensemble strategy to provide a hierarchical feature representation for the hyperspectral data. In the experiments, we select two popular and one challenging hyperspectral data sets for evaluation, and six recently proposed methods are compared. The proposed method achieves respectively 89.55%, 99.36% and 77.90% overall accuracies in the three data sets with 20 training samples per class. The results prove that the performance of the proposed method is superior to some MFF and FE based ones.

Author(s): Lian, Xiaochen | Advisor(s): Yuille, Alan Loddon | Abstract: In this paper, we propose an approach to classify action sequences. We observe that in action sequences the critical features for discriminating between actions occur only within sub-regions of the image. Hence deep network approaches will address the entire image are at a disadvantage. This motivates our strategy which uses static and spatio-temporal visual cues to isolate static and spatio-temporal regions of interest (ROIs). We then use weakly supervised learning to train deep network classifiers using the ROIs as input. More specifically, we combine multiple instance learning (MIL) with convolutional neural networks (CNNs) to select discriminative action cues. This yields classifiers for static images, using the static ROIs, as well as classifiers for short image sequences (16 frames), using spatio-temporal ROIs. Extensive experiments performed on the UCF101 and HMDB51 benchmarks show that both these types of classifiers perform well individually and achieve state of the art performance when combined together. We also show qualitatively that our ROIs (selected by the algorithms) capture the most relevant parts of the image sequences.

This paper reviews some ground breaking development of RF MEMS technology in Australia at the UNSW, over the past decade. It presents some unique and novel designs using RF MEMS switches to achieve reconfigurable RF front-end circuits. These designs include multiport RF MEMS switches, switch matrices, reconfigurable filters and antennas. The resulting devices achieved RF performance that is unmatched by any existing RF andmicrowave technologies.

We expatiate on the design and implementation of multimoding smart home system based on wireless communication protocol of WiFi,BasicRF and infrared ray.We use the MCU CC2530F256 from TI as the main control chip,it reads,transform and revise data acquired by the temperature and humidity sensor and the illuminance sensor,then it delivers the data to the smartphone to display.The system can also learn the functionality of remoters using compression algorithm which can control household appliances,and delivers the learned infrared coding to smartphone to store.The system can use the child node to refile the infrared signal to overcome the block caused by the wall inside home.This smart home product has begun to sell online,after testing and investigation of user feedback,we can see that the system has a high performance and is very stable and reliable,it can accurately display the illuminance,temperature and humidity,and the function of learning infrared signal is also very well,so the system meets the home requirements.