arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2103.04174cs.CVcs.AIcs.LGcs.RO
收藏

用于大规模视频预测的贪婪分层变分自编码器

Greedy Hierarchical Variational Autoencoders for Large-Scale Video Prediction

Stanford University(斯坦福大学)

Bohan Wu, Suraj Nair, Roberto Martin-Martin, Li Fei-Fei, Chelsea Finn

摘要:

一个推广到不同场景的视频预测模型将使智能代理(如机器人)能够通过模型规划来执行各种任务。然而,虽然现有的视频预测模型在小数据集上产生了有希望的结果,但在大型和多样化的数据集上训练时,它们会受到严重的欠拟合。为了解决这一不合适的挑战,我们首先观察到,训练更大视频预测模型的能力通常会受到GPU或TPU内存限制的瓶颈。同时,深层潜变量模型可以通过捕获未来观测的多级随机性来产生更高质量的预测,但这种模型的端到端优化非常困难。我们的关键见解是,分层自动编码器的贪婪和模块化优化可以同时解决内存限制和大规模视频预测的优化挑战。我们介绍了贪婪分层变分自编码器(GHVAE),一种通过对分层自编码器的每一级进行贪婪训练来学习高保真视频预测的方法。与最先进的模型相比,GHVAE在四个视频数据集上的预测性能提高了17-55%,在实际机器人任务上的成功率提高了35-40%,并且可以通过简单地添加更多模块来单调地提高性能。

英文摘要:

A video prediction model that generalizes to diverse scenes would enable intelligent agents such as robots to perform a variety of tasks via planning with the model. However, while existing video prediction models have produced promising results on small datasets, they suffer from severe underfitting when trained on large and diverse datasets. To address this underfitting challenge, we first observe that the ability to train larger video prediction models is often bottlenecked by the memory constraints of GPUs or TPUs. In parallel, deep hierarchical latent variable models can produce higher quality predictions by capturing the multi-level stochasticity of future observations, but end-to-end optimization of such models is notably difficult. Our key insight is that greedy and modular optimization of hierarchical autoencoders can simultaneously address both the memory constraints and the optimization challenges of large-scale video prediction. We introduce Greedy Hierarchical Variational Autoencoders (GHVAEs), a method that learns high-fidelity video predictions by greedily training each level of a hierarchical autoencoder. In comparison to state-of-the-art models, GHVAEs provide 17-55% gains in prediction performance on four video datasets, a 35-40% higher success rate on real robot tasks, and can improve performance monotonically by simply adding more modules.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑