DiscrimNet: Semi-Supervised Action Recognition from Videos using Generative Adversarial Networks
This addresses the problem of action recognition for video analysis, offering an incremental improvement by adapting existing GAN techniques to semi-supervised learning in this domain.
The paper tackles action recognition from videos by proposing a semi-supervised framework using Generative Adversarial Networks (GANs), where a discriminator is pre-trained unsupervisedly and fine-tuned on labeled data, achieving superior or comparable performance to state-of-the-art methods on UCF101 and HMDB51 datasets.
We propose an action recognition framework using Gen- erative Adversarial Networks. Our model involves train- ing a deep convolutional generative adversarial network (DCGAN) using a large video activity dataset without la- bel information. Then we use the trained discriminator from the GAN model as an unsupervised pre-training step and fine-tune the trained discriminator model on a labeled dataset to recognize human activities. We determine good network architectural and hyperparameter settings for us- ing the discriminator from DCGAN as a trained model to learn useful representations for action recognition. Our semi-supervised framework using only appearance infor- mation achieves superior or comparable performance to the current state-of-the-art semi-supervised action recog- nition methods on two challenging video activity datasets: UCF101 and HMDB51.