End-To-End Semi-supervised Learning for Differentiable Particle Filters
This work addresses the challenge of reducing annotated data demand for particle filters in robotics, offering a semi-supervised approach that is incremental over prior methods requiring true states.
The paper tackles the problem of learning dynamic and measurement models for differentiable particle filters without requiring full knowledge of true states, which are often expensive or unavailable, by proposing an end-to-end learning objective based on pseudo-likelihood maximization. The method is evaluated on state estimation tasks in robotics using simulated and real-world datasets, showing improved state estimation when large portions of true states are unknown.
Recent advances in incorporating neural networks into particle filters provide the desired flexibility to apply particle filters in large-scale real-world applications. The dynamic and measurement models in this framework are learnable through the differentiable implementation of particle filters. Past efforts in optimising such models often require the knowledge of true states which can be expensive to obtain or even unavailable in practice. In this paper, in order to reduce the demand for annotated data, we present an end-to-end learning objective based upon the maximisation of a pseudo-likelihood function which can improve the estimation of states when large portion of true states are unknown. We assess performance of the proposed method in state estimation tasks in robotics with simulated and real-world datasets.