Research map / 3D vision / Human pose estimation
Forecasting People Trajectories and Head Poses by Jointly Reasoning on Tracklets and Vislets
Irtiza Hasan, Francesco Setti, Theodore Tsesmelis, Vasileios Belagiannis, Sikandar Amin, Alessio Del Bue, Marco Cristani, Fabio Galasso
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 social pooling (ablation) | MX-LSTM · Zara02 | MAD (m): 0.72 -> 0.35 | Improved |
| add social pooling (ablation) | MX-LSTM · TownCenter | MAD (m): 2.09 -> 1.15 | Improved |
| add view-frustum selection to social pooling (ablation) | MX-LSTM · TownCenter | MAD (m): 1.70 -> 1.15 | Improved |
| add view-frustum selection to social pooling (ablation) | MX-LSTM · Zara02 | MAD (m): 0.36 -> 0.35 | No clear effect |
| model full covariance of tracklet and vislet instead of block-diagonal (ablation) | MX-LSTM · TownCenter | MAD (m): 1.40 -> 1.15 | Improved |
| model full covariance of tracklet and vislet instead of block-diagonal (ablation) | MX-LSTM · Zara01 | MAD (m): 0.60 -> 0.59 | No clear effect |
| use head pose vislets instead of step direction (Pace) (ablation) | MX-LSTM · Zara02 | MAD (m): 0.73 -> 0.35 | Improved |
| use head pose vislets instead of step direction (Pace) (ablation) | MX-LSTM · TownCenter | MAD (m): 1.50 -> 1.15 | Improved |
| use manually annotated head poses at inference instead of estimator (ablation) | MX-LSTM · TownCenter | MAD (m): 1.21 -> 1.15 | Improved |
| train on annotated head poses instead of detector-generated ones (ablation) | MX-LSTM-HPE · Zara02 | MAD (m): 0.51 -> 0.37 | Improved |
| replace Social GAN with MX-LSTM | Social GAN · UCY | MAD (m): 0.65 -> 0.49 | Improved |
| replace Social GAN with MX-LSTM | Social GAN · Zara01 | MAD (m): 0.48 -> 0.59 | Hurt |
Closest papers by meaning
- MX-LSTM: Mixing Tracklets and Vislets to Jointly Forecast Trajectories and Head Poses (CVPR 2018 · similarity 0.96)
- Social LSTM: Human Trajectory Prediction in Crowded Spaces (CVPR 2016 · similarity 0.89)
- STGAT: Modeling Spatial-Temporal Interactions for Human Trajectory Prediction (ICCV 2019 · similarity 0.87)
- Social-Transmotion: Promptable Human Trajectory Prediction (ICLR 2024 · similarity 0.87)
- Multi-modal Knowledge Distillation-based Human Trajectory Forecasting (CVPR 2025 · similarity 0.86)
- A Unified Environmental Network for Pedestrian Trajectory Prediction (AAAI 2024 · 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.
