Most of the existing low-light enhancement research focuses on global illumination enhancement while ignoring the issues of brightness unevenness and color distortion. To address this dilemma, we propose a low-light image enhancement method that can achieve good performance in luminance adjustment and color correction simultaneously. Specifically, the Luminance Adjustment Module is designed to model the global luminance adjustment parameters while taking into account the relationship between global and local illumination features, in order to prevent overexposure or underexposure. Furthermore, we design a Color Correction Module based on the attention mechanism, which utilizes the attention mechanism to capture global color features and correct the color deviation in the illumination-enhanced image. Additionally, we design a color loss function based on a 14-dimensional statistical feature vector related to color, enabling further restoration of the image’s true color. We conduct empirical studies on multiple public low-light datasets, demonstrating that the proposed method outperforms other representative state-of-the-art models regarding illumination enhancement and color correction.

Large Language Model (LLM)-driven social simulation offers a controllable approach for analyzing crisis responses, but existing work on online crises often emphasizes either user engagement prediction or opinion and eWOM evolution. This separation is insufficient for online firestorms, where crisis impact emerges from the coupling between what users express and how participation expands over time. In such events, responsibility attribution, KOL influence, electronic Word of Mouth (eWOM), and user engagement jointly shape collective reactions. To address this gap, we introduce the Crisis Response Interaction Simulation Pipeline (CRISP), a training-free LLM-driven multi-agent framework for online firestorm simulation. CRISP integrates an eWOM Perception module for responsibility attribution and attitude formation with an Engagement Mechanism for predicting participation under evolving KOL influence. Experiments on four heterogeneous Weibo online firestorms across beverage, automobile manufacturing, food service, and education domains show that CRISP reproduces major eWOM and engagement trajectories across different activity scales and interaction structures. Counterfactual interventions on emotional composition and responsibility attribution further produce directionally consistent responses, suggesting mechanism-level validity beyond trajectory fitting. These findings indicate that CRISP provides a framework for analyzing online firestorm evolution and evaluating crisis communication strategies in controllable simulation environments.

In recent years, with the development of computer and networking technology, much more attention has been raised to digital watermark technology. The robustness and invisibility are two main indicators of the digital image watermarking system, and the combination of watermark pre-processing and embedding watermarking algorithm can improve the robustness and reduce error rate. In this paper, the algorithm can extract watermark without the original image, and realize a blind detection of watermark. The experimental results show that this algorithm is good at invisibility, and has good robustness to the attacks of down-sampling, noise, format converting, filtering.

To build the network music big data platform, We need to collect a lot of music resources from the Internet, And store it on big data platforms. Therefore, it is a key problem to build a storage system that is high performance, scalable and capable of supporting big data. The network music data is very large and the metadata structure is not consistent, HBase and other distributed databases are widely used to build it. However, single RowKey index query in HBase is inefficient, it is difficult to meet the demand of big data platform for real-time or quasi-real-time query and statistical analysis. To this end, this paper proposes a solution, First, we migrated to HBase through Sqoop, which stored the music resources stored in multiple relational databases; Then use Phoenix to establish a secondary index query mechanism, build the Phoenix + HBase storage structure, and store management of the heterogeneous music resources; Finally, the readwrite performance and query performance of the storage architecture are tested in the Spark parallel computing framework. Compared with the experimental results, the efficiency of Phoenix + HBase query performance is significantly higher than that of single computer and direct call to HBase, and it can operate unstructured data and structured data, which is in line with the requirement of the storage system of music big data platform.

On the basis of cluster rendering, we present a WAN-based distributed animation rendering system. According to the distribution of resource on WAN and the characteristics of computational nodes, this paper presents how to construct a distributed rendering system on WAN. Combining with P2P technology gives a novel solution to solve the problems such as the performance bottleneck in networks and single point of failure. The system presented in this paper divides the network into multiple clusters according to geographical-proximity principle; each cluster elects a super peer by using Reputation-Aware Selection Algorithm. Hence this method can reduce the communication time of rendering on WAN, improve the robustness of rendering network, and ensure good scalability of the system.