<abstract> <p>Water pollution prevention and control of the Xiang River has become an issue of great concern to China's central and local governments. To further analyze the effects of central and local governmental policies on water pollution prevention and control for the Xiang River, this study performs a big data analysis of 16 water quality parameters from 42 sections of the mainstream and major tributaries of the Xiang River, Hunan Province, China from 2005 to 2016. This study uses an evidential reasoning-based integrated assessment of water quality and principal component analysis, identifying the spatiotemporal changes in the primary pollutants of the Xiang River and exploring the correlations between potentially relevant factors. The analysis showed that a series of environmental protection policies implemented by Hunan Province since 2008 have had a significant and targeted impact on annual water quality pollutants in the mainstream and tributaries. In addition, regional industrial structures and management policies also have had a significant impact on regional water quality. The results showed that, when examining the changes in water quality and the effects of pollution control policies, a big data analysis of water quality monitoring results can accurately reveal the detailed relationships between management policies and water quality changes in the Xiang River. Compared with policy impact evaluation methods primarily based on econometric models, such a big data analysis has its own advantages and disadvantages, effectively complementing the traditional methods of policy impact evaluations. Policy impact evaluations based on big data analysis can further improve the level of refined management by governments and provide a more specific and targeted reference for improving water pollution management policies for the Xiang River.</p> </abstract>

<abstract><p>Deep learning has provided powerful support for person re-identification (person re-id) over the years, and superior performance has been achieved by state-of-the-art. While under practical application scenarios such as public monitoring, the cameras' resolutions are usually 720p, the captured pedestrian areas tend to be closer to $ 128\times 64 $ small pixel size. Research on person re-id at $ 128\times 64 $ small pixel size is limited by less effective pixel information. The frame image qualities are degraded and inter-frame information complementation requires a more careful selection of beneficial frames. Meanwhile, there are various large differences in person images, such as misalignment and image noise, which are harder to distinguish from person information at the small size, and eliminating a specific sub-variance is still not robust enough. The Person Feature Correction and Fusion Network (FCFNet) proposed in this paper introduces three sub-modules, which strive to extract discriminate video-level features from the perspectives of "using complementary valid information between frames" and "correcting large variances of person features". The inter-frame attention mechanism is introduced through frame quality assessment, guiding informative features to dominate the fusion process and generating a preliminary frame quality score to filter low-quality frames. Two other feature correction modules are fitted to optimize the model's ability to perceive information from small-sized images. The experiments on four benchmark datasets confirm the effectiveness of FCFNet.</p></abstract>