It is challenging for the existing single image de-raining algorithms to fully explore the interaction of attention mechanisms in different dimensions. Therefore, an algorithm based on joint attention mechanism was proposed to realize single image de-raining. The algorithm contains a channel attention mechanism and a spatial attention mechanism. Specifically, in the channel attention mechanism, the distribution of rain streak features in each channel was detected and the importance of each feature channel was differentiated. In the spatial attention mechanism, aiming at the spatial relationship of rain streak distribution within channels, the context information was accumulated in a local to global manner to realize efficient and accurate de-raining. Additionally, a deep residual shrinkage network with a soft threshold nonlinear transformation sub-network embedded in the residual module was used to zero out redundant information via a soft threshold function, thereby improving the ability of the CNN in retaining image details in noise. Experiments were carried out on open rainfall data sets and self constructed rainfall data sets. Compared with spatial attention, the joint attention rain removal algorithm improved Peak Signal-to-Noise Ratio (PSNR) by 4.5% and the Structural SIMilarity (SSIM) by 0.3%. Experimental results show that the proposed algorithm can effectively perform single image de-raining and image detail preserving. At the same time, this algorithm outperforms the comparison algorithms in terms of visual effect and quantitative metrics.
As deep residual network has problems such as complex network structure and high time cost in face recognition applications of small mobile devices, a lightweight model based on deep residual network was proposed. Firstly, by simplifying and optimizing the structure of the deep residual network and combining the knowledge transfer method, a lightweight residual network (student network) was reconstructed from the deep residual network (teacher network), which reduced the network structural complexity while ensuring accuracy. Then, in the student network, the parameters of the model were reduced by decomposing standard convolution, thereby reducing the time complexity of the feature extraction network. Experimental results show that on four different datasets such as LFW (Labeled Faces in the Wild), VGG-Face (Visual Geometry Group Face), AgeDB (Age Database) and CFP-FP (Celebrities in Frontal Profile with Frontal-Profile), with the recognition accuracy close to the mainstream face recognition methods, the proposed model has the time of reasoning reaches 16 ms every image, and the speed is increased by 10% to 20%. Therefore, the proposed model can have the speed of reasoning effectively improved with the recognition accuracy basically not reduced.
Sentiment analysis, as a subdivision of Natural Language Processing(NLP), has experienced the development of using sentiment lexicon, machine learning and deep learning to analyze. According to the problem of low accuracy, over fitting phenomenon in training process and low coverage, large workload when compiling the sentiment lexicon when using the generalized deep learning model as a text classifier to analysis of Web text reviews in a specific field, a sentiment analysis model based on sentiment lexicon and stacked residual Bidirectional Long Short-Term Memory (Bi-LSTM) network was proposed. Firstly, the sentiment words in the sentiment lexicon were designed to cover the professional words in the research field of "educational robot", thereby making up for the lack of accuracy of Bi-LSTM model in analyzing such texts. Then, Bi-LSTM and SnowNLP were used to reduce the volume of compilation of the sentiment lexicon. The memory gate and forget gate structures of Long Short-Term Memory (LSTM) network were able to ensure that the relevance of the words before and after in the comment text were fully considered with some analyzed words selected to be forgotten at the same time, thereby avoiding the problem of gradient explosion during the back propagation. After the introduction of the stacked residual Bi-LSTM, not only the number of layers of the model was deepened to 8, but also the "degradation" problem caused by the residual network stacking LSTM was avoided. Finally, by setting and adjusting the score weights of the two parts appropriately, and the sigmoid activation function was used to normalize the total score to the interval of [0,1]. According to the interval division of [0,0.5] and (0.5,1], negative and positive emotions were represented respectively, and sentiment classification was completed. Experimental results show that the sentiment classification accuracy of the proposed classification model for the reviews dataset about "educational robot" is improved by about 4.5 percentage points compared with the standard LSTM model and by about 2.0 percentage points compared with the BERT (Bidirectional Encoder Representation from Transformers). In conclusion, the sentiment classification model based on sentiment lexicon and deep learning classification model was generalized by the proposed model, and by modifying the sentiment words in the lexicon and appropriately adjusting the layer number and the structure of the deep learning model, the proposed model can be applied to accurate sentiment analysis of shopping reviews of all kinds of goods in e-commerce platform, thereby helping enterprises to understand the consumers’ shopping psychology and the market demand, as well as providing consumers with a reference standard for the quality of goods.
Combining Discrete Wavelet Transform (DWT), Stationary Wavelet Transform (SWT) and Non-Local Means (NLM), a new single-frame Super-Resolution (SR) method named DSNLM was proposed to eliminate the blurring effect in wavelet domain SR image. In DSNLM, the subbands were obtained by applying DWT to low-resolution input image, and SWT was simultaneously applied to obtain high frequency subbands; Then NLM filter was applied to these composite subbands along with the interpolated input image. Finally, Inverse Discrete Wavelet Transform (IDWT) was applied to these subbands to obtain the SR image. The experimental and visual results verify the superiority of the proposed method over the conventional image resolution enhancement techniques with improved Peak Signal-to-Noise Ratio (PSNR), Mean Squared Error (MSE) and Structural SIMilarity (SSIM), and it is effective in denoising and blurring.