Robust Autonomy and Decisions Group, The School of Informatics, The University of Edinburgh, 10 Crichton St, Edinburgh EH8 9AB, United Kingdom; The Edinburgh Centre of Robotics, The University of Edinburgh's Bayes Centre, 47 Potterrow, Edinburgh EH8 9BT, United Kingdom; The School of Engineering and Physical Sciences, The Robotarium, Heriot-Watt University, Edinburgh, EH14 4AS, United Kingdom.
Robust Autonomy and Decisions Group, The School of Informatics, The University of Edinburgh, 10 Crichton St, Edinburgh EH8 9AB, United Kingdom; The Edinburgh Centre of Robotics, The University of Edinburgh's Bayes Centre, 47 Potterrow, Edinburgh EH8 9BT, United Kingdom; FiveAI, 5th Floor, Greenside, 12 Blenheim Place, Edinburgh, EH7 5JH, United Kingdom.
Neural Netw. 2020 Dec;132:506-520. doi: 10.1016/j.neunet.2020.09.016. Epub 2020 Sep 26.
This work presents an analysis of the discriminators used in Generative Adversarial Networks (GANs) for Video. We show that unconstrained video discriminator architectures induce a loss surface with high curvature which make optimization difficult. We also show that this curvature becomes more extreme as the maximal kernel dimension of video discriminators increases. With these observations in hand, we propose a methodology for the design of a family of efficient Lower-Dimensional Video Discriminators for GANs (LDVD-GANs). The proposed methodology improves the performance and efficiency of video GAN models it is applied to and demonstrates good performance on complex and diverse datasets such as UCF-101. In particular, we show that LDVDs can double the performance of Temporal-GANs and provide for state-of-the-art performance on a single GPU using the proposed methodology.
这项工作分析了用于视频的生成对抗网络 (GAN) 中的鉴别器。我们表明，无约束的视频鉴别器结构会导致具有高曲率的损失曲面，这使得优化变得困难。我们还表明，随着视频鉴别器的最大核维度的增加，这种曲率变得更加极端。有了这些观察结果，我们提出了一种用于设计一系列用于 GAN 的高效低维视频鉴别器 (LDVD-GAN) 的方法。所提出的方法提高了所应用的视频 GAN 模型的性能和效率，并在复杂多样的数据集（如 UCF-101）上展示了良好的性能。特别是，我们表明 LDVD 可以将 Temporal-GAN 的性能提高一倍，并使用所提出的方法在单个 GPU 上提供最先进的性能。
