arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1810.00746cs.CVcs.LG
收藏

基于综合似然的未来街道场景贝叶斯预测

Bayesian Prediction of Future Street Scenes using Synthetic Likelihoods

Max Planck Institute for Informatics(马克斯·普朗克信息学研究所);Saarland Informatics Campus(萨尔兰信息学园区)

Apratim Bhattacharyya, Mario Fritz, Bernt Schiele

摘要:

对于自主智能体在现实世界中成功运行,预测未来场景状态的能力是一个关键的能力。在现实世界的场景中,未来的状态变得越来越不确定和多模态,特别是在长时间范围内。基于丢弃的贝叶斯推理提供了一种计算上易于处理的,理论上有充分依据的方法来学习可能的假设/模型,以处理不确定的未来,并做出与观察结果相对应的预测--经过很好的校准。然而,事实证明,这样的方法不足以捕捉复杂的现实世界的场景,甚至落后于准确性相比,简单的确定性方法。这是因为使用的对数似然估计不鼓励多样性。在这项工作中,我们提出了一种新的贝叶斯公式预测未来的场景状态,它利用合成似然性,鼓励学习不同的模型,以准确地捕捉未来场景状态的多模态性质。我们表明,我们的方法实现了准确的最先进的预测和校准的概率,通过广泛的实验,对城市景观数据集的场景预测。此外,我们表明,我们的方法概括了不同的任务,如数字生成和降水预测。

英文摘要:

For autonomous agents to successfully operate in the real world, the ability to anticipate future scene states is a key competence. In real-world scenarios, future states become increasingly uncertain and multi-modal, particularly on long time horizons. Dropout based Bayesian inference provides a computationally tractable, theoretically well grounded approach to learn likely hypotheses/models to deal with uncertain futures and make predictions that correspond well to observations -- are well calibrated. However, it turns out that such approaches fall short to capture complex real-world scenes, even falling behind in accuracy when compared to the plain deterministic approaches. This is because the used log-likelihood estimate discourages diversity. In this work, we propose a novel Bayesian formulation for anticipating future scene states which leverages synthetic likelihoods that encourage the learning of diverse models to accurately capture the multi-modal nature of future scene states. We show that our approach achieves accurate state-of-the-art predictions and calibrated probabilities through extensive experiments for scene anticipation on Cityscapes dataset. Moreover, we show that our approach generalizes across diverse tasks such as digit generation and precipitation forecasting.

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。

↑