Learning in an Uncertain World: Representing Ambiguity Through Multiple Hypotheses
This addresses uncertainty in tasks like object detection and pose estimation, offering a principled method that is incremental but broadly applicable.
The paper tackles the problem of uncertainty in prediction tasks by proposing a framework to convert single-prediction models into multiple hypothesis prediction models, showing that these outperform single-hypothesis counterparts across applications like human pose estimation and future prediction.
Many prediction tasks contain uncertainty. In some cases, uncertainty is inherent in the task itself. In future prediction, for example, many distinct outcomes are equally valid. In other cases, uncertainty arises from the way data is labeled. For example, in object detection, many objects of interest often go unlabeled, and in human pose estimation, occluded joints are often labeled with ambiguous values. In this work we focus on a principled approach for handling such scenarios. In particular, we propose a framework for reformulating existing single-prediction models as multiple hypothesis prediction (MHP) models and an associated meta loss and optimization procedure to train them. To demonstrate our approach, we consider four diverse applications: human pose estimation, future prediction, image classification and segmentation. We find that MHP models outperform their single-hypothesis counterparts in all cases, and that MHP models simultaneously expose valuable insights into the variability of predictions.