Abstract
The prominent growth of multimodal systems, which integrate text, speech, vision, and gesture as inputs, has introduced new challenges for software testing. Traditional testing frameworks are not designed to address the dynamic interactions and contextual dependencies inherent to these systems. AI-driven test automation solutions provide transformative solutions by automating test scenario generation, bug detection, and continuous performance monitoring, ensuring efficient testing workflows and integration testing between multiple AI models.
This paper presents a comprehensive review of AI-driven techniques employed for the automated testing of multimodal systems, and critically handling integration of diversified tools, scenario generation frameworks, test data creation approach, and their role in continuous integration pipelines.