Research on Data Desensitization Techniques and Their Applications in Multimodal Environments

In the current era, data has become a critical factor of production, prompting heightened demands for its security. Among various data protection strategies, data desensitization (or data masking) technologies have consistently garnered significant attention from both industry and academia. This paper presents an analysis and comprehensive review of data desensitization techniques and their applications in multimodal environments. First, the developmental trajectory of data desensitization technologies is systematically outlined. A comparative analysis is then conducted to evaluate the advantages and limitations of different desensitization approaches, along with recommendations for their suitable application scenarios. To address the complex demands of data desensitization in multimodal contexts, this study proposes a novel Multimodal Fusion and Multi-layer Collaborative Desensitization Framework. The architecture and core mechanisms of this framework are elaborated in detail. Experimental results demonstrate that the proposed framework outperforms traditional desensitization methods in terms of effectiveness and adaptability.