Cross-dataset palmprint recognition promotes the convenience and flexibility of recognition. However, most current cross-dataset methods need to collect target dataset in advance for model training. They tend overfit this are difficult generalize other unknown datasets. In paper, we propose a novel Palmprint Data Feature Generation (PDFG) method more challenging scenario, Cross-Dataset Recognition with Unseen Target (CDPR-UT). Both data-level feature-level generalization is constructed...

In addition to relying on super computing power and professional domain knowledge, training a high-precision deepfake fingerprint detection model (DFDM) authenticate the fingerprints also requires support of massive private data. To sum up, DFDM should be deemed as intellectual property (IP) trainers, so it is crucial protect IP. Currently, most watermarking-based IP protection schemes are implemented by introducing additional tasks, such constructing trigger sets, fine-tuning weights, etc.,...

Hospital information analysis is a very important way to enhance the medical service. In this paper, hospital system of university implemented for revealing hidden information. First, processing model studied. Then, multi-layer architecture and framework are proposed designed based on thought management system. Third, data statistic mining technologies employed manipulate data, which includes Apriori algorithm, time series so on. Finally, experiments finished show that has intrinsic...

To effectively detect and recognize multicomponent Linear Frequency-Modulated(LFM) emitter signals, a multi-component LFM signal analysis method based on the complex Independent Component Analysis (ICA) which was combined with Fractional Fourier Transform(FRFT) proposed. The idea adopted to this time-domain separation then time-frequency analysis, in low SNR cases.the problem is generally plagued by noised feature extraction of FRFT overcame. Compared traditional analysis,the computer...

Abstract Physical education occupies a very important position in university programs. Cultivating students’ physical fitness and comprehensive ability is the significance of existence development sports. In this paper, after collecting processing data related to tests, K-means clustering algorithm improved cluster analyze different students. Then, FP-growth based on association rules mines relationship between indicators groups students, obtains optimization strategy teaching, introduces...

There are quantities of such sensors as radar, ESM, navigator in aerospace areas and the sequence data is most ordinary sensor domain. How to mine information these has attracted a great interest mining. But easily interfered produces some fluctuant points. When dealing with sequences, traditional similarity measurement Euclidean distance arises large error, especially for unequal length sequence. A novel average weight 1-norm algorithm based on dynamic time warping (DTW) proposed solve this...

In order to advance the assessment capability of significance integrating old-fashioned physical education into school education, an evaluation model founded on big data analysis influencing factors quality is put forward. The was established, and constraint parameter adopted evaluate by method objective elements collection teaching effect. Combined with quantitative statistical structural parameters system, important classical institution realized, inversion feature quantity extracted....

Abstract The mine Roofbolter is widely used in drilling and supporting of underground roof side wall, which greatly alleviates the imbalance driving anchor. However, with progress sensor technology, electrical electronic electro-hydraulic proportional control technology continuous improvement users’ requirements for safety efficiency bolt support, domestic foreign mining equipment manufacturers have carried out development automatic anchor cable support equipment, especially to improve...

A large amount of data has been accumulated. with the development Internet industry. Many problems have exposed explosion: 1. The contradiction between privacy and collaborations; 2. ownership right usage; 3. legality collection 4. relationship governance rules; 5. Traceability evidence chain. In order to face such a complicated situation, many algorithms were proposed developed. This article tries build model from perspective blockchain make some breakthroughs.Internet Of Rights(IOR) uses...