Jiangbo Shu

Adaptive total variation-based spectral deconvolution with the split Bregman method.

Spectroscopic data often suffer from common problems of band overlap and noise. This paper presents a maximum a posteriori (MAP)-based algorithm for the band overlap problem. In the MAP framework, the likelihood probability density function (PDF) is constructed with Gaussian noise assumed, and the prior PDF is constructed with adaptive total variation (ATV) regularization. The split Bregman iteration algorithm is employed to optimize the ATV spectral deconvolution model and accelerate the speed of the spectral deconvolution. The main advantage of this algorithm is that it can obtain peak structure information as well as suppress noise simultaneity. Simulated and real spectra experiments manifest that this algorithm can satisfactorily recover the overlap peaks as well as suppress noise and are robust to the regularization parameter.

Blind spectrum reconstruction algorithm with L0-sparse representation

Raman spectrum often suffers from band overlap and Poisson noise. This paper presents a new blind Poissonian Raman spectrum reconstruction method, which incorporates the L0-sparse prior together with the total variation constraint into the maximum a posteriori framework. Furthermore, the greedy analysis pursuit algorithm is adopted to solve the L0-based minimization problem. Simulated and real spectrum experimental results show that the proposed method can effectively preserve spectral structure and suppress noise. The reconstructed Raman spectra are easily used for interpreting unknown chemical mixtures.

Blind image restoration with sparse priori regularization for passive millimeter-wave images

A blind image restoration method for the passive millimeter-wave images is proposed.The regularization item is constructed as the hyper-Laplace function ||x||0.6.A data-selected matrix is proposed to estimate the accurate pint spread function.The proposed method improves the resolution of the PMMW image. Passive millimeter wave imaging often suffers from issues such as low resolution, noise, and blurring. In this study, a blind image restoration method for the passive millimeter-wave images (PMMW) is proposed. The purpose of the proposed method is to simultaneously solve the point spread function (PSF) and restoration image. In this method, the data fidelity item is constructed based on Gaussian noise assuming, and the regularization item is constructed as the hyper-Laplace function ||x||0.6, which is fitted according to the high-resolution PMMW images. Moreover, a data-selected matrix is proposed to select the regions that are helpful for estimating the accurate PSF. The proposed method has been applied to simulated and real PMMW image experiments. Comparative results demonstrate that the proposed method significantly outperforms the state-of-the-art deblurring methods on both qualitative and quantitative assessments. The proposed method improves the resolution of the PMMW image and makes it more preferable for object recognition.

Exploration of Hybrid Teaching of Software Engineering on StarC

In view of the problem of the low efficiency in traditional classroom teaching due to the limitation in time and space, an exploration which combines real classroom with virtual classroom in hybrid learning was proposed. We chose the teaching of a software engineering course and used starC as the teaching support tool for analysis. In our study, the teaching process was divided into several teaching units, and each teaching unit was further divided into several activity units. The content was organized in the form of topicalities, where students are allowed to choose the learning content according to their study plans and preferences. Through the questionnaire survey which includes the indicators of participation and satisfaction among the students on both traditional learning and hybrid learning, it is found that the students on hybrid learning have higher participation and satisfaction than that on traditional learning. This indicated that hybrid learning could effectively improve teaching effectiveness.

A content-based recommendation algorithm for learning resources

Automatic multimedia learning resources recommendation has become an increasingly relevant problem: it allows students to discover new learning resources that match their tastes, and enables the e-learning system to target the learning resources to the right students. In this paper, we propose a content-based recommendation algorithm based on convolutional neural network (CNN). The CNN can be used to predict the latent factors from the text information of the multimedia resources. To train the CNN, its input and output should first be solved. For its input, the language model is used. For its output, we propose the latent factor model, which is regularized by L1-norm. Furthermore, the split Bregman iteration method is introduced to solve the model. The major novelty of the proposed recommendation algorithm is that the text information is used directly to make the content-based recommendation without tagging. Experimental results on public databases in terms of quantitative assessment show significant improvements over conventional methods. In addition, the split Bregman iteration method which is introduced to solve the model can greatly improve the training efficiency.

Automatic Recommendation Technology for Learning Resources with Convolutional Neural Network

Automatic learning resources recommendation has become an increasingly relevant problem: it allows students to discover new learning resources that matches their tastes, and enables e-learning system to target their learning resources to the right students. In this paper, we propose an automatic learning resources recommendation algorithm based on convolutional neural network (CNN). The CNN can be used to predict the latent factors from the text information. To train the CNN, its input and output should be solved firstly. For its input, the language model is employed. For its output, we propose the latent factor model, which is regularized by L1-norm. Furthermore, the split Bregman iteration method is introduced to solve the model. The major novelty of the proposed recommendation algorithm is that a new CNN is constructed to make personalized recommendations. Experimental results on public database in terms of quantitative assessment show significant improvements over conventional methods. Especially, it can also work well when the existing recommendation algorithms suffer from the cold-start problem.

Expanded autoencoder recommendation framework and its application in movie recommendation

Automatic recommendation has become a popular research field: it allows the user to discover items that match their tastes. In this paper, we proposed an expanded autoencoder recommendation framework. The stacked autoencoders model is employed to extract the feature of input then reconstitution the input to do the recommendation. Then the side information of items and users is blended in the framework and the Huber function based regularization is used to improve the recommendation performance. The proposed recommendation framework is applied on the movie recommendation. Experimental results on a public database in terms of quantitative assessment show significant improvements over conventional methods.

Blind infrared spectroscopic data restoration with the similarity of multi-scales regularization

Band overlap and random noise exist widely when the spectra are captured using an infrared spectrometer, especially when the problems of instrument aging has become more and more serious recently. In this paper, via introducing the similarity of multiscales, a blind spectral deconvolution method is proposed. Considering similarity of the latent spectrum between different scales, it is used as a prior to constrain the estimated latent spectrum similar to pre-scale to reduce artifacts which is produced from deconvolution. Experiments indicate that the proposed method is able to obtain better performance than the state-of-the-art methods, and obtain satisfying deconvolution results with fewer artifacts. The recovered infrared spectra can easily extract the spectral features and recognize the unknown objects.

Parametric spectral signal restoration via maximum entropy constraint and its application

In this paper, we will propose a new framework which can estimate the desired signal and the instrument response function (IRF) simultaneously from the degraded spectral signal. Firstly, the spectral signal is considered as a distribution, thus, new entropy (called differential-entropy, DE) is defined to measure the distribution with a uniform distribution, which allows negative value existing. Moreover, the IRF is parametrically modeled as a Lorentzian function. Comparative results manifest that the proposed method outperforms the conventional methods on peak narrowing and noise suppression. The deconvolution IR spectrum is more convenient for extracting the spectral feature and interpreting the unknown chemical mixtures.

Interactive Visualization and Its Application in Junior Middle School Chemistry Teaching

With the development of interactive visualization technology, it is an excellent choice to apply it to the field of education. In this paper, we first briefly introduce interactive visualization and design principles. Then, we select participants from junior high school in Wuhan (N=55) and use a between-subjects design with participants assigned randomly to one of two instructional conditions: interactive chemistry experiment platform (ICEP) and PowerPoint. The results show that the students who use the ICEP have better performance in learning and applying knowledge compared with traditional teaching. Finally, we propose the research direction of interactive visualization in education field and hope to provide a reference for the teachers and students to carry out research.