Aiming at the problem that the classification and localization sub-tasks in object detection require large receptive field and high resolution respectively， and it is difficult to achieve a balance between these two contradictory requirements， a feature pyramid network algorithm based on attention mechanism for object detection was proposed. In the algorithm， multiple different receptive fields were integrated to obtain richer semantic information， multi-scale feature maps were fused in the way of paying more attention to the importance of different feature maps， and the fused feature maps were further refined under the guidance of the attention mechanism. Firstly， multi-scale receptive fields were obtained through multiple atrous convolutions with different dilation rates， which enhanced the semantic information with the preservation of the resolution. Secondly， through the Multi-Level Fusion （MLF）， multiple feature maps of different scales were fused after changing to the same resolution through upsampling or pooling operations. Finally， the proposed Attention-guided Feature Refinement Module （AFRM） was used to refine the fused feature maps to enhance semantic information and eliminate the aliasing effect caused by fusion. After replacing the Feature Pyramid Network （FPN） in Faster R-CNN with the proposed feature pyramid， experiments were performed on MS COCO 2017 dataset. The results show that when the backbone network is ResNet （Residual Network） with a depth of 50 and 101， with the use of the proposed algorithm， the Average Precision （AP） of the model reaches 39.2% and 41.0% respectively， which is 1.4 and 1.0 percentage points higher than that of Faster R-CNN using the original FPN， respectively. It can be seen that the proposed feature pyramid network algorithm can replace the original feature pyramid to be better applied in the object detection scenarios.

Current skill teaching methods of manipulator mainly construct a virtual space through three-dimensional reconstruction technology for manipulator to simulate and train. However， due to the different visual angles between human and manipulator， the traditional visual information reconstruction methods have large reconstruction errors， long time， and need harsh experimental environment and many sensors， so that the skills learned by manipulator in virtual space can not be well transferred to the real environment. To solve the above problems， a visual information real-time reconstruction method for manipulator operation was proposed. Firstly， information was extracted from real-time RGB images through Mask-Region Convolutional Neural Network（Mask-RCNN）. Then， the extracted RGB images and other visual information were jointly encoded， and the visual information was mapped to the three-dimensional position information of the manipulator operation space through Residual Neural Network-18 （ResNet-18）. Finally， an outlier adjustment method based on Cluster Center DIStance constrained （CC-DIS） was proposed to reduce the reconstruction error， and the adjusted position information was visualized by Open Graphics Library （OpenGL）. In this way， the three-dimensional real-time reconstruction of the manipulator operation space was completed. Experimental results show that the proposed method has high reconstruction speed and reconstruction accuracy. It only takes 62.92 milliseconds to complete a three-dimensional reconstruction with a reconstruction speed of up to 16 frames per second and a reconstruction relative error of about 5.23%. Therefore， it can be effectively applied to the manipulator skill teaching tasks.

To address the issues of blurred contours and lost details of portrait image with motion blur after restoration， a moving portrait deblurring method based on multi-level jump residual group Generation Adversarial Network （GAN） was proposed. Firstly， the residual block was improved to construct the multi-level jump residual group module， and the structure of PatchGAN was also improved to make GAN better combine with the image features of each layer. Secondly， the multi-loss fusion method was adopted to optimize the network to enhance the real texture of the reconstructed image. Finally， the end-to-end mode was used to perform blind deblurring on the motion blurred portrait image and output clear portrait image. Experimental results on CelebA dataset show that the Peak Signal-to-Noise Ratio （PSNR） and Structural SIMilarity （SSIM） of the proposed method are at least 0.46 dB and 0.05 higher than those of the Convolutional Neural Network （CNN）-based methods such as DeblurGAN （Deblur GAN）， Scale-Recurrent Network （SRN） and MSRAN （Multi-Scale Recurrent Attention Network）. At the same time， the proposed method has fewer model parameters， faster restoration， and more texture details in the restored portrait images.

Collaborative Filtering （CF） algorithm can realize personalized recommendation on the basis of the similarity between items or users. However， data sparsity has always been one of the challenges faced by CF algorithm. In order to improve the prediction accuracy， a CF algorithm based on Collaborative Training and Boosting （CFCTB） was proposed to solve the problem of sparse user-item scores. First， two CFs were integrated into a framework by using collaborative training， pseudo-labeled samples with high confidence were added to each other’s training set by the two CFs， and Boosting weighted training data were used to assist the collaborative training. Then， the weighted integration was used to predict the final user scores， and the accumulation of noise generated by pseudo-labeled samples was avoided effectively， thereby further improving the recommendation performance. Experimental results show that the accuracy of the proposed algorithm is better than that of the single models on four open datasets. On CiaoDVD dataset with the highest sparsity， compared with Global and Local Kernels for recommender systems （GLocal-K）， the proposed algorithm has the Mean Absolute Error （MAE） reduced by 4.737%. Compared with ECoRec （Ensemble of Co-trained Recommenders） algorithm， the proposed algorithm has the Root Mean Squared Error （RMSE） decreased by 7.421%. The above rasults verify the effectiveness of the proposed algorithm.

Concerning the ranging inaccuracy problem based on radio signal phase information under multi-path environments, a two-step ranging approach based on double tags was proposed. Each target was attached with double tags. Through single frequency subcarrier amplitude modulation, firstly, the wrapped phase information of carrier signal was extracted, the distance between reader and tag within half wavelength of carrier signal was calculated and fine ranging estimation value was achieved. Secondly, the unwrapped phase information of subcarrier signal was extracted, and the integral multiple of half wavelength within the distance of reader and tag was calculated. Thirdly, the average multiple was calculated between double tags, the distance of average multiple of half wavelength was used as coarse ranging value. Finally, the final ranging result was estimated by the sum of the fine ranging value and coarse ranging value. Additionally, single reader and double-tag based geometric localization method was introduced to reduce the cost of hardware facilities. The simulation results show that, under multi-path environments, compared with the directly ranging with subcarrier phase, the average ranging error of double tags based two-step ranging approach is reduced by 35%, and the final average localization error is about 0.43 m, and the maximum error is about 1 m. The proposed approach can effectively improve the accuracy of phase based localization technology and also reduce the hardware cost.

Conventional Direction-Of-Arrival (DOA) estimation approaches suffer from low angular resolution or relying on a large number of snapshots. The sparsity-based SPICE can work with few snapshots and has high resolution and low sidelobe level, but it only applies to narrowband signals. To solve the above problems, a new FrFT-SPICE method was proposed to estimate the DOA of wideband chirp signals with high resolution based on a few snapshots. First, the wideband chirp signal was taken on the Fractional Fourier Transform (FrFT) under a specific order so that the chirp wave in time domain could be converted into sine wave with single frequency in FrFT domain. Then, the steering vector of the received signal was obtained in FrFT domain. Finally, SPICE algorithm was utilized with the obtained steering vector to estimate the DOA of the wideband chirp. In the simulation with the same scanning grid and same snapshots, the DOA resolution level of the proposed FrFT-SPICE method was better than that of the FrFT-MUSIC method which combines MUltiple SIgnal Classification (MUSIC) algorithm and FrFT algorithm; and compared to the SR-IAA which utilizes Spatial Resampling (SR) and IAA (Iterative Adaptive Approach), the proposed method had a better accuracy. The simulation results show that the proposed method can estimate the DOA of wideband chirp signals with high accuracy and resolution based on only a few snapshots.

Based on the theory of Restless Multi-Armed Bandit (RMAB) model, a novel mechanism of dynamic spectrum access was proposed for the problem that how to coordinate multiple user access multiple idle channels. Firstly, concerning the channel sensing error of the cognitive user being existed in the practical network, the Whittle index policy which can deal with sensing error effectively was derived. In this policy, the users achieved one belief value for every channel based on the historical experience accumulation and chose the channel, which was needed to sense and access, by considering the immediate and future rewards based on the belief values. Secondly, this paper used the multi-bid auction algorithm to deal with the collision among secondary users when they selected the channels to improve the spectrum utilization. The simulation results demonstrate that, in the same environment, the cognitive users with the proposed mechanism have higher throughtput than the mechanism without dealing with sensing error or without multi-bid.

Currently, to develop context-aware applications that are flexible and adaptable is complex and laborious. There are many unexpected cases to handle. As one of the important approaches to achieve context-awareness, predicate detection can represent context effectively. However, how predicate detection supports the development of context-aware applications on a real device is still largely unknown. In order to cope with these issues, a simple scenario was created. Predicate detection was practically applied to control the car running in a designated environment. The original context was formally modeled and contextual properties were specified into snapshot predicates and sequence predicates. By detecting these specified predicates in the case study, predicate detection was applied to the robot car. The performance analysis shows that predicate detection can effectively detect the car's contextual properties and successfully help the car finish the running task.

Articulatory features, which represent the articulatory information, can help prosodic features to improve the performance of tone recognition. In this paper, a set of 19 pronunciation categories was given according to the pronunciation characteristics of initials and finals. Besides, 19 articulatory tandem features, which are the posteriors of speech signal belonging to the 19 pronunciation categories, were obtained by hierarchical multilayer perceptron classifiers. Then these articulatory tandem features, as well as prosodic features, were used for tone modeling. Tone recognition experiments of three kinds of tone models indicate that about 5% absolute increase of accuracy can be achieved when using both articulatory features and prosodic features. When the proposed tone model is integrated into LVSCR (Large Vocabulary Continuous Speech Recognition) system, the character error rate is reduced significantly.

In order to enhance the performance of the synchronization and demodulation, a Very high frequency (VHF) Data Broadcast (VDB) signal demodulation algorithm based on the solution of differential equation was proposed. This algorithm eliminated the synchronization performance deterioration caused by the frequency offset. And frame synchronization, bit synchronization, frequency offset estimation and correction could be completed within a single set of synchronization symbols. The simulation results show that the method is effective to enhance the VDB signal demodulation performance.

This paper researched the fast networking Media Access Control (MAC) of mobile sensor networks based on unslotted CSMA/CA (Carrier Sense Multiple Access with Collision Avoidance) random access algorithm. This paper analyzed and optimized the networking process by preassigning the channels, simplified the association process using direct transmission instead of indirect transmission, and improved the CSMA/CA mechanism in order to reduce the collision. The simulation and field test results show that the proposed method can reduce the collision and the networking time effectively compared with the IEEE 802.15.4 protocol.

To remove the stripe noise from the line scan images, a transform domain destriping method which combined Fourier transform and wavelet decomposition was proposed. Firstly, the image was decomposed using multi-resolution wavelet decomposition to separate the subband which contained the stripe noise from other subbands. Then the subband that contained stripe noise was transformed into Fourier coefficients. The Fourier coefficients were processed by a band-stop filter to remove the stripe noise. The live collected cotton foreign fiber images with stripe noise were used in the simulation experiment. The experimental results indicate that the proposed approach which combined Fourier transform with wavelet decomposition can effectively remove the stripe noise from the image while preserving the characteristics of the original image. It gets better destriping effect than just using Fourier transform or wavelet decomposition separately.

Concerning the changeability of the speech signal caused by co-articulation phenomenon in Chinese speech recognition, a syllable-based acoustic modeling method was proposed. Firstly, context independent syllable-based acoustic models were trained, and the models were initialized by intra-syllable IFs based diphones to solve the problem of training data sparsity. Secondly, the inter-syllable co-articulation effect was captured by incorporating inter-syllable transition models into the recognition system. The experiments conducted on “863-test” dataset show that the relative character error rate is reduced by 12.13%. This proves that syllable-based acoustic model and inter-syllable transition model are effective in solving co-articulation effect.

By analyzing the current rate control algorithm, i.e., JCTVC-H0213 for high efficiency video coding, a new fast and effective method to derive the quantization parameter was proposed, relying on the ratio of the targeted bitrate to the actual coding bitrate of the previous frame. The proposed algorithm was implemented on the HM6.3 platform in the constant bit rate case. The experimental results show that the proposed algorithm outperforms JCTVC-H0213 rate control algorithm. A performance gain in variety of Peak Signal-to-Noise Ratio (PSNR) of more than 2.4 dB can be observed. In conclusion, the proposed algorithm achieves outstanding rate distortion performance with a low complexity, which makes it well suited to real-time scenarios.

The evolutionary mode of Differential Evolution (DE) was analyzed, and modified differentiation operator and selection operator with self-accelerated characteristic were proposed. Then the Self-Accelerated and Variable Neighbourhood searching of Differential Evolution (SAVNDE) algorithm was advanced using these new operators and variable neighbourhood search which improved the local search ability of algorithm. On the basis of the three evolution models, the simulation results on five classical benchmark functions show that SAVNDE has the same convergence rate of DE, and can achieve more optimization results in shorter time.

The original New Edge-Directed Interpolation (NEDI) algorithm is of high complexity, difficult for hardware implementation, and the interpolated images may suffer from blurring edges around edge area. To achieve a better subjective quality, an improved NEDI algorithm was proposed in this paper. In the new algorithm, a circular window was adopted, and the interpolation coefficient calculation was calculated only once, which could be reused in interpolating the center-pixels, thus the errors introduced by iterative computation were avoided and the interpolation time was saved. As to non-center pixels, six original neighbors were involved to estimate local covariance characteristics at high resolution. In comparison with the results of bi-cubic interpolation and the traditional NEDI, the experimental results indicate that proposed algorithm can eliminate the sawtooth of the interpolated picture in large-scale, and decrease the computational complexity.

Recently, Lin, Tang and Wang proposed a multi-prime RSA based on a star architecture of key distribution and made use of it to construct a centralized group key management scheme. According to several main security requirements of group key management, from the perspective of cryptographic engineering practice and applying computational number theory, four kinds of attacks against this scheme were proposed: a ring idempotent attack, a chosen plaintext attack，an attack of extracting high order integer roots, and a collusion attack based on the elliptic curve factoring method and Chinese remainder theorem. The mathematical analysis and cryptanalysis indicate that under certain conditions these attacks can be realized efficiently, and it is the characteristic of "without rekeying the key server's encryption exponent" that causes such security risks.

According to the development of Web field, the construction of Web application based on J2EE was introduced. The popular approach of deploying the static and dynamic content on to Application Server was discussed, and the original approach of dividing files between the Web server and Application Server was researched, furthermore a new approch of performance improvement of Web application by separating static and dynamic content is advanced. Finally, using IBM HTTP Server and WebSphere Application Server, the efficiency comparison between the two approach was presented.