MissFormer: (In-)Attention-Based Handling of Missing Observations for Trajectory Filtering and Prediction

Abstract

In applications such as object tracking, time-series data inevitably carry missing observations. Following the success of deep learningbased models for various sequence learning tasks, these models increasingly replace classic approaches in object tracking applications for inferring the objects' motion states. While traditional tracking approaches can deal with missing observations, most of their deep counterparts are, by default, not suited for this. Towards this end, this paper introduces a transformer -based approach for handling missing observations in variable input length trajectory data. The model is formed indirectly by successively increasing the complexity of the demanded inference tasks. Starting from reproducing noise-free trajectories, the model then learns to infer trajectories from noisy inputs. By providing missing tokens, binary-encoded missing events, the model learns to in-attend to missing data and infers a complete trajectory conditioned on the remaining inputs. In the case of a sequence of successive missing events, the model then acts as a pure prediction model. The abilities of the approach are demonstrated on synthetic data and real-world data reflecting prototypical object tracking scenarios.

Contrastive Masked Transformers for Forecasting Renal Transplant Function

“…Considering the use of the transformer models, the attention mechanism showed promising results in missing data imputation for structural [27] and trajectory data [2,9]. In particular, the attention mask was used to investigate the robustness of a vanilla encoder-decoder transformer and a Bidirectional Transformer (BERT) model [7] while missing 1 to 6 point's coordinates out of 32 for forecasting the people trajectories.…”