This paper explores the generative adversarial network-based error concealing algorithm in panoramic video to further enhance the transmission quality. Specifically, a new network model is proposed for error concealing in panoramic video, which consists primarily of a generator network and two discriminator networks. Among them, the generator network is used to complement the missing parts of lossy frames during video transmission, and the discriminator networks contain a global discriminator network and a local discriminator network, which jointly distinguish the pictures generated by the generator network from the real pictures. Furthermore, this paper also proposes a new mask set. The panoramic video is divided into relatively moving area and relatively stationary area according to its characteristics. For the missing information in relatively moving area, it is mainly repaired by the model proposed in this paper, while for the relatively stationary area, the corresponding part of previous frame is used to achieve error concealment. By comparing the performance of proposed method with that of common GANs, the effectiveness of proposed model is verified through the results of subjective and objective aspects, respectively.