Human Action Generation with Generative Adversarial Networks
This work addresses the need for generating novel human actions for practical applications, but it appears incremental as it builds on existing GAN and autoencoder methods without a major breakthrough.
The paper tackles the problem of generating novel sequences of human motions by introducing a model combining an autoencoder and a GAN, conditioned on initial states and class labels, and demonstrates its ability to produce diverse action styles on the NTU RGB+D dataset.
Inspired by the recent advances in generative models, we introduce a human action generation model in order to generate a consecutive sequence of human motions to formulate novel actions. We propose a framework of an autoencoder and a generative adversarial network (GAN) to produce multiple and consecutive human actions conditioned on the initial state and the given class label. The proposed model is trained in an end-to-end fashion, where the autoencoder is jointly trained with the GAN. The model is trained on the NTU RGB+D dataset and we show that the proposed model can generate different styles of actions. Moreover, the model can successfully generate a sequence of novel actions given different action labels as conditions. The conventional human action prediction and generation models lack those features, which are essential for practical applications.