Research map / Multi-agent systems and games / Human-AI and goal-conditioned
Collaborative multi-agent perception
Shares features over communication graphs among agents for efficient collaborative perception and ad hoc teamwork.
86 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2017–2026) · an approach in Human-AI and goal-conditioned, keyword cluster “communication · ai · collaboration”
State History Tried Baselines Checklist Preprints Who Export
Where it stands
Not enough benchmark data Fewer than four papers report the same benchmark, so there is no record to follow.
24 papers in 2023–24 and 45 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic grew, from 8.1% to 11.7%.
Open questions the data points to
- Papers report V2X-ViT on OPV2V (AP0.7) anywhere from 66.42 to 75.7: comparisons against it depend on whose number you take.
How this is decided
A benchmark is tracked when at least four papers of this approach report their own method on the same dataset and metric in at least two different years. Its record is the best number reported up to 2024 compared with the best reported in 2025–26: still improving means at least 1% better, flat means at most 0.2% better with at least two papers from 2025–26 trying. Numbers on a different scale from the rest (a fraction among percentages) and far outliers are left out. Results come from different papers with different setups, so they are signals, not head-to-head comparisons. The share of papers is counted at ICML, NeurIPS, ICLR, CVPR and AAAI.
How it got here
Most cited paper of each year
- 2017 · Collaborative Deep Reinforcement Learning for Joint Object Search (CVPR 2017 · 88 citations)
- 2018 · Gaussian Process Decentralized Data Fusion Meets Transfer Learning in Large-Scale Distributed Cooperative Perception (AAAI 2018 · 13 citations)
- 2019 · Emergent Coordination Through Competition (ICLR 2019 · 81 citations)
- 2020 · When2com: Multi-Agent Perception via Communication Graph Grouping (CVPR 2020 · 273 citations)
- 2021 · Trajectory Diversity for Zero-Shot Coordination (ICML 2021 · 18 citations)
- 2022 · Where2comm: Communication-Efficient Collaborative Perception via Spatial Confidence Maps (NeurIPS 2022 · 96 citations)
- 2023 · How2comm: Communication-Efficient and Collaboration-Pragmatic Multi-Agent Perception (NeurIPS 2023 · 20 citations)
- 2024 · Communication-Efficient Collaborative Perception via Information Filling with Codebook (CVPR 2024 · 52 citations)
- 2025 · CoSDH: Communication-Efficient Collaborative Perception via Supply-Demand Awareness and Intermediate-Late Hybridization (CVPR 2025 · 22 citations)
- 2026 · InfoCom: Kilobyte-Scale Communication-Efficient Collaborative Perception with Information Bottleneck (AAAI 2026 · 4 citations)
Earliest papers here
- Collaborative Deep Reinforcement Learning for Joint Object Search (CVPR 2017 · 88 citations)
- Collaborative Planning with Encoding of Users' High-Level Strategies (AAAI 2017 · 35 citations)
- Gaussian Process Decentralized Data Fusion Meets Transfer Learning in Large-Scale Distributed Cooperative Perception (AAAI 2018 · 13 citations)
What has been tried
57 papers here report 563 measured changes (a component added, removed or swapped, compared with a baseline): helped 71%, no clear effect 7%, hurt 17%, mixed 4%. Papers mostly publish changes that helped, so "helped" is over-represented.
Ablations in recent papers
| What was changed | Measured on | Result | Effect | Paper |
|---|---|---|---|---|
| use code-index representation instead of transmitting feature vectors | Ablation setting in Fig. 9a · communication cost | 208 times | Helped | Communication-Efficient Collaborative Perception via Information Filling with Codebook (CVPR 2024 · 52 citations) |
| Add MEM | V2XSim · AP@0.7 | 64.28 -> 67.64 | Helped | Multi-agent Collaborative Perception via Motion-aware Robust Communication Network (CVPR 2024 · 37 citations) |
| Add attention-based feature fusion (AFF) | V2V4Real, noise 0.4/0.4 · AP@0.5 | 51.71 -> 62.24 | Helped | ERMVP: Communication-Efficient and Collaboration-Robust Multi-Vehicle Perception in Challenging Environments (CVPR 2024 · 33 citations) |
| Add confidence-aware late fusion | OPV2V · AP@0.7 | 92.60% -> 92.92% | Helped | CoSDH: Communication-Efficient Collaborative Perception via Supply-Demand Awareness and Intermediate-Late Hybridization (CVPR 2025 · 22 citations) |
| add knowledge distillation | V2X-Sim 1.0 · AP at IoU 0.5 | 1.41% | Helped | What Makes Good Collaborative Views? Contrastive Mutual Information Maximization for Multi-Agent Perception (AAAI 2024 · 12 citations) |
| remove the LLM-based reward function | TDW-MAT w/ Oracle Perception · Average transport rate (%) | 93 -> 89−4.30% | Hurt | Collaborative Tree Search for Enhancing Embodied Multi-Agent Collaboration (CVPR 2025 · 6 citations) |
| Adversarial diversity objectives (BRDiv, LIPO) with any constant uniform alpha | Repeated Matrix Game and Weighted Cooperative Reaching · whether MCS members are preferred by diversity objective | Hurt | Minimum Coverage Sets for Training Robust Ad Hoc Teamwork Agents (AAAI 2024 · 6 citations) | |
| add Collaboration Adapter and Agent Prompt | DAIR-V2X (component ablation) · AP@70 | 0.217 -> 0.418 | Helped | CoPEFT: Fast Adaptation Framework for Multi-Agent Collaborative Perception with Parameter-Efficient Fine-Tuning (AAAI 2025 · 5 citations) |
| Add atomic-action intention sharing | Home service, α = 1 · TS | 161 -> 159.4 | Helped | MRBTP: Efficient Multi-Robot Behavior Tree Planning and Collaboration (AAAI 2025 · 5 citations) |
| remove channel selection module | OPV2V pp8-pp4 · AP@0.7 | 60.9 -> 59.9 | Hurt | One is Plenty: A Polymorphic Feature Interpreter for Immutable Heterogeneous Collaborative Perception (CVPR 2025 · 5 citations) |
| Vary the number of subteams, using five subteams or the optimal k-means silhouette score | Real-world football scenarios · threat reduction | Helped | Optimising Spatial Teamwork Under Uncertainty (AAAI 2025 · 2 citations) | |
| Add the multiscale pyramid | OPV2V-H heterogeneous collaboration · AP70 | 0.578 -> 0.628 | Helped | An Extensible Framework for Open Heterogeneous Collaborative Perception (ICLR 2024 · 2 citations) |
Baseline numbers that disagree
The same baseline on the same benchmark and metric, reported more than 5% apart by different papers here. Training budgets, data splits and reimplementations differ, so state which number you compare with.
| Baseline | Benchmark | Metric | Reported | Papers |
|---|---|---|---|---|
| V2X-ViT | OPV2V | AP0.7 | 66.42 – 75.7 | 2 papers, e.g. NeurIPS 2023 and CVPR 2024 |
What reviewers will expect
Datasets and baselines named in the 28 papers of this approach from 2025–2026 that have a full-text summary. Leaving out the common ones is a frequent review question.
Datasets
- OPV2V · 4 of 28 papers
- DAIR-V2X · 4 of 28 papers
- V2X-Sim · 2 of 28 papers
- OPV2V-H · 2 of 28 papers
- V2V4Real · 2 of 28 papers
Baselines
- Late Fusion · 3 of 28 papers
- MPDA · 3 of 28 papers
- Random · 2 of 28 papers
- V2X-ViT · 2 of 28 papers
- PnPDA · 2 of 28 papers
- IPPO · 2 of 28 papers
- PBT · 2 of 28 papers
- STAMP · 2 of 28 papers
Recent preprints
Loading recent arXiv preprints…
Who publishes on it
Institutions on the most papers of this approach since 2023 (of 69 papers; from OpenAlex and the papers' affiliations).
- Shanghai Jiao Tong University · 5 papers
- Google · 4 papers
- National University of Singapore · 3 papers
- Nanyang Technological University · 3 papers
- Beijing University of Posts and Telecommunications · 3 papers
- Tencent · 3 papers
- Massachusetts Institute of Technology · 2 papers
- University of Texas at Austin · 2 papers
- City University of Hong Kong · 2 papers
- Wuhan University of Technology · 2 papers
Use it
Download BibTeX of 34 key papersOpen the topic in the interactive map
Get alerts. Describe your idea in scime atlas and save it (free scime account): you get a weekly email when new preprints land close to it.
Related approaches
Same keyword cluster
- Human-AI coordination agents (125 papers)
- Boundedly rational agent modeling (87 papers)
- Theory-of-mind social reasoning (69 papers)
- Emergent communication in referential games (50 papers)
Elsewhere in the topic
- World-model-based agents (110 papers)
- Imitation and skill learning (100 papers)
- Language-grounded reinforcement agents (96 papers)
- Embodied agent learning (93 papers)
- Intrinsically motivated exploration (86 papers)
- Interactive multi-agent trajectory prediction (75 papers)
- Learning-guided MAPF search (73 papers)
- Vision-language navigation agents (66 papers)
Benchmark numbers, ablations and baseline numbers are extracted by AI from each paper's full text and can be wrong; the paper is the authority. Data: results index of 2026-11-02, citation counts from OpenAlex (2026-10-09).