Ambiguity in Sequential Data: Predicting Uncertain Futures With Recurrent Models

Ambiguity is inherently present in many machine learning tasks, but especially for sequential models seldom accounted for, as most only output a single prediction. In this work we propose an extension of the Multiple Hypothesis Prediction (MHP) model to handle ambiguous predictions with sequential data, which is of special importance, as often multiple futures are equally likely. Our approach can be applied to the most common recurrent architectures and can be used with any loss function. Additionally, we introduce a novel metric for ambiguous problems, which is better suited to account for uncertainties and coincides with our intuitive understanding of correctness in the presence of multiple labels. We test our method on several experiments and across diverse tasks dealing with time series data, such as trajectory forecasting and maneuver prediction, achieving promising results.

I. Introduction

Ambiguity and uncertainty are inherently present in many machine learning tasks, both of sequential and non-sequential nature. A vehicle approaching an intersection might turn left or right, while in text generation (used, e.g., in mobile phones for auto-completion) multiple characters or words might be equally likely to follow the current one. Fig. 1 visualizes the multiple trajectories possible when encountering a roundabout.