Toward Reliable Human Pose Forecasting with Uncertainty
This work addresses the lack of unified evaluation and uncertainty analysis in human pose forecasting, which is crucial for applications like robotics and human-computer interaction, though it is incremental in advancing existing methods.
The paper tackles the problem of unreliable human pose forecasting by developing an open-source library with standardized benchmarks and introducing methods to model aleatoric and epistemic uncertainty, resulting in up to 25% improvements in short-horizon forecasting on datasets like Human3.6M without loss on longer horizons.
Recently, there has been an arms race of pose forecasting methods aimed at solving the spatio-temporal task of predicting a sequence of future 3D poses of a person given a sequence of past observed ones. However, the lack of unified benchmarks and limited uncertainty analysis have hindered progress in the field. To address this, we first develop an open-source library for human pose forecasting, including multiple models, supporting several datasets, and employing standardized evaluation metrics, with the aim of promoting research and moving toward a unified and consistent evaluation. Second, we devise two types of uncertainty in the problem to increase performance and convey better trust: 1) we propose a method for modeling aleatoric uncertainty by using uncertainty priors to inject knowledge about the pattern of uncertainty. This focuses the capacity of the model in the direction of more meaningful supervision while reducing the number of learned parameters and improving stability; 2) we introduce a novel approach for quantifying the epistemic uncertainty of any model through clustering and measuring the entropy of its assignments. Our experiments demonstrate up to $25\%$ improvements in forecasting at short horizons, with no loss on longer horizons on Human3.6M, AMSS, and 3DPW datasets, and better performance in uncertainty estimation. The code is available online at https://github.com/vita-epfl/UnPOSed.