Research map / Learning theory and algorithms / Matrix and tensor methods
A Geometric Framework for Statistical Analysis of Trajectories With Distinct Temporal Spans
Rudrasis Chakraborty, Vikas Singh, Nagesh Adluru, Baba C. Vemuri
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 |
|---|---|---|---|
| represent trajectories as subspaces on a hypersphere instead of approximating each by a geodesic | Trajectory statistics framework · OASIS C vs. ND group test | approximated p-value: 0.38 -> 0.035 | Improved |
| represent trajectories as subspaces on a hypersphere instead of approximating each by a geodesic | Trajectory statistics framework · OASIS C vs. D group test | approximated p-value: 0.53 -> 0.065 | Improved |
| represent trajectories as subspaces on a hypersphere instead of approximating each by a geodesic | Trajectory statistics framework · OASIS D vs. ND group test | approximated p-value: 0.46 -> 0.051 | Improved |
| use the hypersphere framework for PG features | Linear SVM on principal geodesics · OASIS C vs. ND | accuracy (%): 78.95 -> 94.73 | Improved |
| use the hypersphere framework for PG features | Linear SVM on principal geodesics · OASIS C vs. D | accuracy (%): 83.33 -> 88.89 | Improved |
| use the hypersphere framework for PG features | Linear SVM on principal geodesics · OASIS D vs. ND | accuracy (%): 86.96 -> 91.31 | Improved |
| use the hypersphere framework for PG features | Linear SVM on principal geodesics · OASIS C vs. D | sensitivity: 0.86 -> 0.86 | No clear effect |
| use the hypersphere framework instead of geodesic approximation for regression | MVKR regression · OASIS regression | R2 statistic: 0.37 -> 0.64 | Improved |
| use the hypersphere framework instead of geodesic approximation for regression | MVKR regression · HCP regression, ProcSpeed_Unadj | R2 statistic: 0.02 -> 0.78 | Improved |
| use the hypersphere framework instead of geodesic approximation for regression | MVKR regression · HCP regression, ListSort_Unadj | R2 statistic: 0.16 -> 0.75 | Improved |
| use the hypersphere framework instead of geodesic approximation on synthetic geodesic trajectories | Frechet mean of trajectories · Synthetic geodesics on S2 | mean trajectory (qualitative) | No clear effect |
| use PGA on the hypersphere representation instead of space-time features or [25] | Gymnastics video summarization · Gymnastics (5 activities, 8 gymnasts each) | class separation in 2D biplot (qualitative) | Improved |
Closest papers by meaning
- Learned Trajectory Embedding for Subspace Clustering (CVPR 2024 · similarity 0.84)
- Geometric Analysis of Nonlinear Manifold Clustering (NeurIPS 2024 · similarity 0.82)
- GeoSPRINT: Geometric Redundancy-Aware Step Pruning for Inference in Diffusion Trajectories (NeurIPS 2026 · similarity 0.82)
- Homogeneous Dynamics Space for Heterogeneous Humans (CVPR 2025 · similarity 0.82)
- Sports-Traj: A Unified Trajectory Generation Model for Multi-Agent Movement in Sports (ICLR 2025 · similarity 0.82)
- Geometric Trajectory Diffusion Models (NeurIPS 2024 · similarity 0.82)
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.
