Research map / Probabilistic and scientific modeling / Molecules and proteins
FoSS: Modeling Long-Range Dependencies and Multimodal Uncertainty in Trajectory Prediction via Fourier-State Space Integration
Yizhou Huang, Genze Jiang, Yihua Cheng, Kezhi Wang
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 |
|---|---|---|---|
| add the frequency-domain branch (ablation) | FoSS · Argoverse 2 (ablation) | minADE: 0.71 -> 0.65 | Improved |
| add the frequency-domain branch (ablation) | FoSS · Argoverse 2 (ablation) | minFDE: 1.36 -> 1.29 | Improved |
| add progressive helix reordering (ablation) | FoSS · Argoverse 2 (ablation) | minADE: 0.69 -> 0.65 | Improved |
| add Fourier selective SSM (ablation) | FoSS · Argoverse 2 (ablation) | minADE / minFDE / MR: 0.70/1.35/0.17 -> 0.65/1.29/0.15 | Improved |
| use cross-attention fusion instead of Concat+MLP (ablation) | FoSS · Argoverse 2 (ablation) | minADE / minFDE / MR: 0.69/1.33/0.16 -> 0.65/1.29/0.15 | Improved |
| replace the Mamba time-domain branch with a Transformer (ablation) | FoSS · Argoverse 2 | minADE: 0.61 -> 0.75 | Hurt |
| replace the Mamba time-domain branch with an LSTM (ablation) | FoSS · Argoverse 2 | minFDE: 1.07 -> 1.28 | Hurt |
| FoSS versus DenseTNT on b-minFDE | FoSS · Argoverse 2 | b-minFDE: 1.98 -> 1.69 | Improved |
| FoSS versus QCNet on parameters | FoSS · Argoverse 2 | Params (M): 7.71 -> 4.18 | Improved |
| FoSS versus QCNet on FLOPs | FoSS · Argoverse 2 | GFLOPs: 45.3 -> 22.1 | Improved |
| FoSS versus QCNet on latency | FoSS · Argoverse 2 | ms: 71 -> 64 | Improved |
| FoSS versus DeMo on minADE | FoSS · Argoverse 1 | minADE_1: 1.65 -> 1.67 | Hurt |
Closest papers by meaning
- Trajectory Mamba: Efficient Attention-Mamba Forecasting Model Based on Selective SSM (CVPR 2025 · similarity 0.87)
- C2F-TP: A Coarse-to-Fine Denoising Framework for Uncertainty-Aware Trajectory Prediction (AAAI 2025 · similarity 0.86)
- UniMotion: A Unified Motion Framework for Simulation, Prediction and Planning (NeurIPS 2025 · similarity 0.86)
- ModeSeq: Taming Sparse Multimodal Motion Prediction with Sequential Mode Modeling (CVPR 2025 · similarity 0.85)
- DeMo: Decoupling Motion Forecasting into Directional Intentions and Dynamic States (NeurIPS 2024 · similarity 0.85)
- OneVL: One-Step Latent Reasoning and Planning with Vision-Language Explanation (NeurIPS 2026 · similarity 0.85)
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.
