AI-based Text to Scene Generation as Part of a Pandemic Compliant Infrastructure

After the 2020–2022 COVID-19 pandemic, it is observed that a responsive medical infrastructure and damage control techniques, including application of technology, have become more important than ever before. Many such technologies, including the Internet of Things (IoT), and artificial intelligence (AI)-aided decision-making have become relevant. Within such a framework, text-to-scene generation through the incorporation of Artificial Intelligence (AI) becomes a critical element of a pandemic-compliant infrastructure design. Our suggested architecture uses AI algorithms to create sceneries that are dynamically adjusted for pandemic compliance. These sceneries cover a wide range of settings, including as public locations, medical institutions, virtual meeting rooms, and more. Our approach makes use of generator and discriminator blocks as part of a Generative Adversarial Network (GAN) which is further enhanced by several attention layers to generate scenes from test descriptions. The output of the proposed approach ensures that scene creation regarding safety precautions like mask wearing, social distance, and sanitization procedures are included in generated output.