DiscrimNet: Semi-Supervised Action Recognition from Videos using Generative Adversarial Networks

Abstract

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.

References in corpus (14)

• Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
• UCF101: A Dataset of 101 Human Actions Classes From Videos in The Wild
• Empirical Evaluation of Rectified Activations in Convolutional Network
• NIPS 2016 Tutorial: Generative Adversarial Networks
• InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets
• Generating Videos with Scene Dynamics
• Adversarial Feature Learning
• Semi-Supervised Learning with Generative Adversarial Networks
• Deep multi-scale video prediction beyond mean square error
• Unsupervised and Semi-supervised Learning with Categorical Generative Adversarial Networks
• Deep Networks with Stochastic Depth
• Unsupervised Learning of Visual Representations using Videos
• Fusing Multi-Stream Deep Networks for Video Classification
• Self-Supervised Video Representation Learning With Odd-One-Out Networks

Cited by in corpus (6)

• Rethinking Monocular Depth Estimation with Adversarial Training
• VideoSSL: Semi-Supervised Learning for Video Classification
• Iterate & Cluster: Iterative Semi-Supervised Action Recognition
• Multi-Level Sequence GAN for Group Activity Recognition
• Predicting the Future: A Jointly Learnt Model for Action Anticipation
• Fine-grained Action Segmentation using the Semi-Supervised Action GAN