针对多智能体感知不确定性，提出细粒度模态级协作方法
Active Asymmetric Multi-Agent Multimodal Learning under Uncertainty
- 按模态建模不确定性，动态选择可靠传感器组合
- 在自动驾驶事故检测中提升18.7%识别率
- 适合传感器异构、部分失效的复杂场景
多智能体系统日益配备异构多模态传感器，虽增强感知能力，却引入模态特异性与智能体依赖的不确定性。现有协作框架通常在智能体层面推理，假设感知同质且隐式处理不确定性，难以应对传感器损坏。我们提出主动异构多智能体多模态不确定性学习（A2MAML），一种面向不确定性的模态级协作方法。A2MAML将每种模态特征建模为带不确定性预测的随机估计，主动选择可靠智能体-模态组合，并通过贝叶斯反方差加权融合信息。该方法实现细粒度模态级融合，支持异构模态可用性，并提供抑制受损或噪声模态的理论机制。在连通自动驾驶场景下的协同事故检测实验表明，A2MAML持续优于单智能体及协作基线，事故检测率最高提升18.7%。
原文摘要 · Abstract (English)
Multi-agent systems are increasingly equipped with heterogeneous multimodal sensors, enabling richer perception but introducing modality-specific and agent-dependent uncertainty. Existing multi-agent collaboration frameworks typically reason at the agent level, assume homogeneous sensing, and handle uncertainty implicitly, limiting robustness under sensor corruption. We propose Active Asymmetric Multi-Agent Multimodal Learning under Uncertainty (A2MAML), a principled approach for uncertainty-aware, modality-level collaboration. A2MAML models each modality-specific feature as a stochastic estimate with uncertainty prediction, actively selects reliable agent-modality pairs, and aggregates information via Bayesian inverse-variance weighting. This formulation enables fine-grained, modality-level fusion, supports asymmetric modality availability, and provides a principled mechanism to suppress corrupted or noisy modalities. Extensive experiments on connected autonomous driving scenarios for collaborative accident detection demonstrate that A2MAML consistently outperforms both single-agent and collaborative baselines, achieving up to 18.7% higher accident detection rate.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。