As a creative method for virtual human individuals based on multiple fusion technologies such as artificial intelligence, computer graphics, and speech synthesis, virtual human technology has developed rapidly since its birth, and continuous discussions and studies have been conducted in both academia and industry. Starting from the film and television industries, the cross-disciplinary application of virtual human has been continuously recognized and applied in fields such as media, games, and finance. Although virtual human has achieved sufficient development and innovation, it faces many challenges such as emotion recognition, privacy, and security, as well as the uncanny valley effect. This article starts with the development history of virtual human and analyzes the current academic research status and application scenarios in combination with the characteristics, technical architecture, and application of virtual human technology. At the same time, this article sorts out seven mainstream application scenarios of virtual human and analyzes their main advantages and possible future challenges. This article provides a valuable reference for subsequent related research by exploring development trends, application fields, and future research trends in virtual human.