Abstract: Benefiting from the development of the deep generative model, face manipulation technology has made great breakthroughs. The deep face forgery technology, represented by Deepfake, has rapidly become popular on the Internet and received extensive attention from academia and industry. This deep forgery technique synthesizes fake face videos by exchanging the identity information between original and target faces or by editing the attribute information of target faces. The deep face forgery technique has inspired many related entertainment applications, such as using facial substitution to replace the user's face in a movie clip, or using expression reenactment to drive a static portrait of a famous person. However, the current deep face forgery technology is still in the fast-growing stage, and its authenticity and naturality require further improvement. On the other hand, this kind of deep face forgery can easily be used by malicious criminals to produce pornographic movies, fake news, or even political rumors about important dignitaries, which is a great potential threat to national security and social stability, so the defense technology of face forgery videos is crucial. In order to reduce the negative impact of deep face forgery videos, many researchers have investigated the detection and identification techniques of face forgery videos, proposing a series of defense methods from different perspectives. Unfortunately, due to the single form of dataset distribution, inconsistent evaluation metrics and lack of initiative, etc., there is still a long way to go before the detection technology can be applied practically. In fact, the research on deep face forgery and detection techniques remains in the developmental stage, their connotations and extensions are being rapidly updated and iterated. In this review, we will make a scientific and systematic summary of the main research work so far, and briefly analyze the limitations of existing technologies. Finally, we will discuss the potential challenges and development directions of deep face forgery and detection technology, in order to provide insights for future research work in the field.