Research map / Vision-language and multimodal / Cross-modal learning
CooperScene: Multi-Modal Cooperative Autonomy Benchmark with C-V2X Communication Characterization
Bo Wu, Ruoshen Mo, Justin Yue, Yanyu Zhang, Janice Nguyen, Guoyuan Wu, Amit Roy-Chowdhury, Matthew Barth, Hang Qiu
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 |
|---|---|---|---|
| limit sharing to real C-V2X bandwidth instead of unlimited network | CoSDH V+2V+I · CooperScene detection | mAP@0.3: 0.81 -> 0.75 | Hurt |
| limit sharing to real C-V2X bandwidth instead of unlimited network | V2VNet V+2V+I · CooperScene detection | mAP@0.3: 0.81 -> 0.32 | Hurt |
| add more cooperating agents under C-V2X | V2VNet · CooperScene detection | mAP@0.3: 0.43 (V+I) -> 0.32 (V+2V+I) | Hurt |
| add more cooperating agents under C-V2X | CoSDH · CooperScene detection | mAP@0.3: 0.57 (V+I) -> 0.75 (V+2V+I) | Improved |
| add more cooperating agents under unlimited network | CoBEVT · CooperScene detection | mAP@0.3: 0.52 (V+I) -> 0.82 (V+2V+I) | Improved |
| use communication-efficient feature selection (CoSDH) instead of full feature sharing (V2VNet) | CoSDH · CooperScene detection V+2V+I | sharing size (MB): 30.0 -> 0.1785 | Improved |
| add camera to LiDAR for cooperative detection (ablation) | BEVFusion V+2V · CooperScene detection BEV | mAP@0.7: 0.56 -> 0.61 | Improved |
| add camera to LiDAR for single vehicle (ablation) | BEVFusion V · CooperScene detection BEV | mAP@0.3: 0.78 -> 0.80 | Improved |
| add cooperating vehicles for motion prediction | CMP · CooperScene motion | minADE@5s: 1.4022 -> 1.2893 | Improved |
| add peer vehicles for short-horizon prediction | V2VNet · CooperScene motion | minADE@1s: 0.4895 -> 0.7610 | Hurt |
| add one peer vehicle for CMP | CMP · CooperScene motion | minADE@5s: 1.4022 -> 1.6416 | Hurt |
| use spatial-temporal ICP instead of no ICP or pairwise ICP (ablation) | spatial alignment · CooperScene | RMSE (m): 0.2 m for most frames | Improved |
Closest papers by meaning
- End-to-End Autonomous Driving Through V2X Cooperation (AAAI 2025 · similarity 0.89)
- V2X-Real: a Largs-Scale Dataset for Vehicle-to-Everything Cooperative Perception (ECCV 2024 · similarity 0.88)
- VIPS: Vehicle-Infrastructure Cooperative Planning Benchmark via Pseudo-Simulation (ECCV 2026 · similarity 0.87)
- UrbanIng-V2X: A Large-Scale Multi-Vehicle, Multi-Infrastructure Dataset Across Multiple Intersections for Cooperative Perception (NeurIPS 2025 · similarity 0.84)
- CoSDH: Communication-Efficient Collaborative Perception via Supply-Demand Awareness and Intermediate-Late Hybridization (CVPR 2025 · similarity 0.84)
- Autonomous Driving Research Requires a Community-Driven Data Paradigm (NeurIPS 2026 · similarity 0.83)
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.
