Research map / Multi-agent systems and games / Human-AI and goal-conditioned
When2com: Multi-Agent Perception via Communication Graph Grouping
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use learned handshake communication to select when and whom to communicate with (ablation) | When2com · MRMS | Avg. mIoU | Helped |
| Remove handshake communication (ablation) | When2com · Multi-agent 3D shape recognition | Accuracy | Hurt |
| Increase query size from 1 to 128 while keeping key size at 16 (ablation) | When2com · Multi-agent 3D shape recognition | Grouping accuracy | Helped |
| Increase key size (ablation) | When2com · Multi-agent 3D shape recognition | Grouping accuracy | Helped |
| Use query size 4 (ablation) | When2com · Multi-agent 3D shape recognition | Classification accuracy | Helped |
| Use When2com instead of TarMac | Multi-agent 3D shape recognition · Degraded split | Accuracy (%): 78.73 -> 80.72 | Helped |
| Use When2com instead of TarMac | MRMS · MRMS | Avg. mIoU: 53.92 -> 57.28 | Helped |
| Use When2com instead of TarMac | MRMPS · MRMPS | Avg. mIoU: 40.59 -> 40.82 | Helped |
| Use When2com instead of TarMac | MRMS · MRMS | Bandwidth (Mbpf): 2.5 -> 0.385 | Helped |
| Use When2com instead of TarMac | MRMPS · MRMPS | Bandwidth (Mbpf): 2.0 -> 0.55 | Helped |
| Use When2com instead of TarMac | Multi-agent 3D shape recognition · Multi-agent 3D shape recognition | Bandwidth (MBpf): 8 -> 1.32 | Mixed |
| Use query messages of increasing size (ablation) | When2com · Multi-agent 3D shape recognition | Classification accuracy | Helped |
Closest papers by meaning
- Communication-Efficient Collaborative Perception via Information Filling with Codebook (CVPR 2024 · similarity 0.89)
- An Extensible Framework for Open Heterogeneous Collaborative Perception (ICLR 2024 · similarity 0.88)
- CoSDH: Communication-Efficient Collaborative Perception via Supply-Demand Awareness and Intermediate-Late Hybridization (CVPR 2025 · similarity 0.88)
- CoPEFT: Fast Adaptation Framework for Multi-Agent Collaborative Perception with Parameter-Efficient Fine-Tuning (AAAI 2025 · similarity 0.88)
- Pragmatic Heterogeneous Collaborative Perception via Generative Communication Mechanism (NeurIPS 2025 · similarity 0.87)
- Learning Multi-Agent Communication from Graph Modeling Perspective (ICLR 2024 · similarity 0.87)
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.