The light field camera has significantly advanced conventional imaging methods and microscopy over the past decades, providing high-dimensional information in 2D images and enabling a variety of applications. However, inherent shortcomings persist, mainly due to the complex optical setup and the trade-off between resolution. In this work, we propose a Neural Defocus Light Field (NDLF) rendering method, which constructs the light field without a micro-lens array but achieves the same resolution as the original image. The basic unit of NDLF is the 3D point spread function (3D-PSF), which extends the 2D-PSF by incorporating the focus depth axis. NDLF can directly solve the distribution of PSFs in 3D space, enabling direct manipulation of the PSF in 3D and enhancing our understanding of the defocus process. NDLF achieves the focused images rendering by redefining the focus images as slices of the NDLF, which are superpositions of cross-sections of the 3D-PSFs. NDLF modulates the 3D-PSFs using three multilayer perceptron modules, corresponding to three Gaussian-based models from coarse to fine. NDLF is trained on 20 highresolution (1024 × 1024) images at different focus depths, enabling it to render focused images at any given focus depth. The structural similarity index between the predicted and measured focused images is 0.9794. Moreover, we developed a hardware system to collect the high resolution focused images with corresponding focus depth, and depth maps. NDLF achieves high-resolution light field imaging with a single-lens camera and also resolves the distribution of 3D-PSFs in 3D space, paving the way for novel lightfield synthesis techniques and deeper insights into defocus blur.

Sentiment analysis is significant for excavating text opinion. There are two issues in the foreign exchange (Forex) field. 1) In sentiment orientation, most researches focus on product reviews, lack fine-grained sentiment analysis for Forex news. 2) In sentiment intensity, most works consider the intensity of sentiment words but ignore the significance of field characteristics. Aiming at the two problems, a fine-grained Sentiment Analysis model (shorted as WD-SA) is established, which integrates with the Weight of sentiment words and Domain features. First, the semantic information of text is embedded into a vector based on word2vec. Then, sentiment orientation is detected by a method, which combines machine learning algorithm and the weight of sentiment words. Finally, features are extracted to investigate the intensity of news. The experimental results show that our algorithm outperforms the state-of-the-art.

The forecast and analysis of the trend of the RMB exchange rate have been deeply explored by many researchers in the financial field, but the combination of public opinion sentiment data and historical market data to forecast the RMB exchange rate in the short term has been studied little. With the rapid development of the Internet, the influence of public opinion sentiment on the economy and society is increasing. Online public opinion sentiment data not only have an impact on stock prices [1] and commodity prices, but also have a significant impact on foreign exchange (Forex) rates. However, the public opinion data are not applied for the RMB exchange rate forecast, because the impact of public events on the exchange rate is ignored. Besides, the lack of exact temporal sliding window of public opinion ignores its timeliness and sensibility.

Recursive neural network (RvNN) has been proved to be an effective and promising tool to learn sentence representations by explicitly exploiting the sentence structure. However, most existing work can only exploit simple tree structure, e.g., binary trees, or ignore the order of nodes, which yields suboptimal performance. In this paper, we proposed a novel neural network, namely TreeNet, to capture sentences structurally over the raw unconstrained constituency trees, where the number of child nodes can be arbitrary. In TreeNet, each node is learning from its left sibling and right child in a bottom-up left-to-right order, thus enabling the net to learn over any tree. Furthermore, multiple soft gates and a memory cell are employed in implementing the TreeNet to determine to what extent it should learn, remember and output, which proves to be a simple and efficient mechanism for semantic synthesis. Moreover, TreeNet significantly suppresses convolutional neural networks (CNN) and Long Short-Term Memory (LSTM) with fewer parameters. It improves the classification accuracy by 2%-5% with 42% of the best CNN’s parameters or 94% of standard LSTM’s. Extensive experiments demonstrate TreeNet achieves the state-of-the-art performance on all four typical text classification tasks.

The aim of scene semantic segmentation is to label each pixel with a class which it belongs to in high level cognition. State-of-art works mainly adapt convolutional neural networks originally designed for image classification to make dense prediction. However the inner structure of scene itself and its stuff is more flexible and variable, which is distinct from the objects in image classification task. Therefore we propose a gate function based structure-aware convolution for deep neural networks with the ability of modeling inner variance in scene. The gate function is a RNN-based learnable function or a handcrafted one, which is applied to distinguish efficient activations from convolution area. It is proved that dilated convolution is a subclass of gate function. As shown in our experiments on scene datasets, the proposed convolution method efficiently improves the accuracy of current semantic segmentation systems by partly replacing original networks' convolution layers with ours.