Research map / Multi-agent systems and games / Human-AI and goal-conditioned
TacSIm: A Dataset and Benchmark for Football Tactical Style Imitation
Peng Wen, Yuting Wang, Qiurui 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 |
|---|---|---|---|
| extend prediction horizon from 3.0s to 10.0s (ablation) | BC · TacSIm, 150 cells | imitation score (%): 37.86 -> 8.63 | Hurt |
| extend prediction horizon from 3.0s to 10.0s (ablation) | CoDAIL · TacSIm, 150 cells | imitation score (%): 50.89 -> 28.37 | Hurt |
| extend prediction horizon from 3.0s to 10.0s (ablation) | CMIL · TacSIm, 60 cells | imitation score (%): 46.06 -> 27.57 | Hurt |
| medium grid (150 cells) instead of finest grid (1768 cells) (ablation) | CoDAIL · TacSIm, 3.0s | imitation score (%): 27.10 -> 50.89 | Improved |
| medium grid (150 cells) instead of coarse grid (60 cells) (ablation) | CoDAIL · TacSIm, 3.0s | imitation score (%): 46.63 -> 50.89 | Improved |
| medium grid (240 cells) instead of 150 cells (ablation) | CMIL · TacSIm, 3.0s | imitation score (%): 42.98 -> 47.87 | Improved |
| coordinated multi-agent modeling (CMIL) instead of independent BC | BC · TacSIm, 60 cells, 3.0s | imitation score (%): 37.05 -> 46.06 | Improved |
| IRL instead of BC | BC · TacSIm, 60 cells, 3.0s | imitation score (%): 37.05 -> 34.18 | No clear effect |
| diffusion-based DRAIL vs CoDAIL for long horizon | CoDAIL · TacSIm, 10.0s, 60 cells | imitation score (%): 33.00 -> 28.65 | Hurt |
Closest papers by meaning
- Event2Tracking: Reconstructing Multi-Agent Soccer Trajectories Using Long-Term Multimodal Context (AAAI 2025 · similarity 0.84)
- SportR: A Benchmark for Multimodal Large Language Model Reasoning in Sports (ICLR 2026 · similarity 0.84)
- Towards Universal Soccer Video Understanding (CVPR 2025 · similarity 0.82)
- Sports-Traj: A Unified Trajectory Generation Model for Multi-Agent Movement in Sports (ICLR 2025 · similarity 0.82)
- LATTE-MV: Learning to Anticipate Table Tennis Hits from Monocular Videos (CVPR 2025 · similarity 0.81)
- SoccerMaster: A Vision Foundation Model for Soccer Understanding (CVPR 2026 · similarity 0.81)
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.
