Ruiz Adria, Rudovic Ognjen Oggi, Binefa Xavier, Pantic Maja
IEEE Trans Image Process. 2018 Apr 25. doi: 10.1109/TIP.2018.2830189.
We propose a Multi-Instance-Learning (MIL) approach for weakly-supervised learning problems, where a training set is formed by bags (sets of feature vectors or instances) and only labels at bag-level are provided. Specifically, we consider the Multi-Instance Dynamic-Ordinal-Regression (MI-DOR) setting, where the instance labels are naturally represented as ordinal variables and bags are structured as temporal sequences. To this end, we propose Multi-Instance Dynamic Ordinal Random Fields (MI-DORF). In this framework, we treat instance-labels as temporally-dependent latent variables in an Undirected Graphical Model. Different MIL assumptions are modelled via newly introduced high-order potentials relating bag and instance-labels within the energy function of the model. We also extend our framework to address the Partially-Observed MI-DOR problems, where a subset of instance labels are available during training.We show on the tasks of weakly-supervised facial behavior analysis, Facial Action Unit (DISFA dataset) and Pain (UNBC dataset) Intensity estimation, that the proposed framework outperforms alternative learning approaches. Furthermore, we show that MIDORF can be employed to reduce the data annotation efforts in this context by large-scale.
我们提出了一种用于弱监督学习问题的多实例学习（MIL）方法，其中训练集由包（特征向量或实例集）组成，并且仅提供包级别的标签。具体来说，我们考虑多实例动态序数回归（MI-DOR）设置，其中实例标签自然地表示为序数变量，并且包被构造为时间序列。为此，我们提出了多实例动态序数随机场（MI-DORF）。在这个框架中，我们将实例标签视为无向图形模型中随时间变化的潜在变量。通过在模型的能量函数中引入新的高阶势来对不同的MIL假设进行建模，这些高阶势将包和实例标签联系起来。我们还扩展了我们的框架以解决部分观察到的MI-DOR问题，即在训练期间只有一部分实例标签可用的情况。我们在弱监督面部行为分析、面部动作单元（DISFA数据集）和疼痛（UNBC数据集）强度估计任务上表明，所提出的框架优于其他学习方法。此外，我们表明在这种情况下，MIDORF可以大规模减少数据标注工作。
