Research map / Probabilistic and scientific modeling / Variational and generative sampling
Differentiable and Stable Long-Range Tracking of Multiple Posterior Modes
Ali Younis, Erik B. Sudderth
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| truncate gradients at each resampling step (TG-MDPF) instead of IWSG (ablation) | MDPF · House3D | NLL: 8.18 -> 8.38 | Hurt |
| truncate gradients at each resampling step (TG-MDPF) (ablation) | MDPF · House3D | RMSE: 30.59 -> 35.77 | Hurt |
| decouple resampling and posterior mixtures with separate weights and bandwidths (A-MDPF) | MDPF · House3D | NLL: 8.18 -> 8.09 | Improved |
| A-MDPF vs MDPF on RMSE | MDPF · House3D | RMSE: 30.59 -> 30.51 | No clear effect |
| use MDPF with IWSG instead of OT-PF | MDPF · House3D | RMSE: 159.57 -> 30.59 | Improved |
| use MDPF with IWSG instead of TG-PF | MDPF · House3D | NLL: 15.50 -> 8.18 | Improved |
| use MDPF instead of LSTM | MDPF · House3D | RMSE: 48.51 -> 30.59 | Improved |
| replace IWSG with implicit reparameterization gradients (IRG-MDPF) (ablation) | MDPF · Bearings-only tracking | training stability | Hurt |
| truncated gradients (TG-MDPF) instead of IWSG (ablation) | MDPF · Bearings-only tracking | accuracy | Hurt |
| Epanechnikov kernel replacing Gaussian in KDE (ablation) | MDPF · Bearings-only tracking | stability across 11 runs | No clear effect |
| IWSG vs truncated gradients for learning parameters (ablation) | discriminative PF · synthetic bimodal linear-Gaussian | parameter convergence | Improved |
| IRG vs IWSG for parameter learning (ablation) | discriminative PF · synthetic bimodal linear-Gaussian | parameter convergence | Hurt |
Closest papers by meaning
- Learning to be Smooth: An End-to-End Differentiable Particle Smoother (NeurIPS 2024 · similarity 0.91)
- Stochastic Localization via Iterative Posterior Sampling (ICML 2024 · similarity 0.85)
- Nonlinear Filtering with Brenier Optimal Transport Maps (ICML 2024 · similarity 0.84)
- Motion2VecSets: 4D Latent Vector Set Diffusion for Non-rigid Shape Reconstruction and Tracking (CVPR 2024 · similarity 0.84)
- Learning 3D Particle-based Simulators from RGB-D Videos (ICLR 2024 · similarity 0.84)
- Partially Observed Trajectory Inference using Optimal Transport and a Dynamics Prior (ICLR 2025 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
