Differentiable and Stable Long-Range Tracking of Multiple Posterior Modes

arXiv:2404.08789v1 Announce Type: new
Abstract: Particle filters flexibly represent multiple posterior modes nonparametrically, via a collection of weighted samples, but have classically been applied to tracking problems with known dynamics and observation likelihoods. Such generative models may be inaccurate or unavailable for high-dimensional observations like images. We instead leverage training data to discriminatively learn particle-based representations of uncertainty in latent object states, conditioned on arbitrary observations via deep neural network encoders. While prior discriminative particle filters have used heuristic relaxations …

abstract arxiv collection cs.ai cs.lg cs.ro data differentiable dynamics filters generative generative models images multiple observation particle posterior samples tracking training training data type via