用深度学习与强化学习结合，让声呐自动找最佳视角识别水下垃圾。
A Hybrid Model-Based and Model-Free Framework for Active Multi-View Viewpoint Optimization in Sonar Target Recognition
- 融合卷积网络与强化学习，自动选择最优观测视角。
- 相比传统方法减少感知步骤和运动成本，识别准确率相当。
- 适合水下机器人实时目标识别，无需复杂计算或标注角度。
本文提出一种基于模型与无模型相结合的主动多视角目标识别框架，用于前视声呐系统。采用卷积神经网络（CNN）提供数据驱动的观测似然，通过基于Radon的朝向估计实现无需角度标注的视角感知。训练阶段，信息增益奖励引导近端策略优化（PPO）代理学习一个考虑信念的状态决策策略；部署时，仅依赖CNN进行信念更新，即可实时完成视角选择，避免了耗时的在线部分可观测马尔可夫决策过程（POMDP）搜索。在海洋垃圾前视声呐数据集上的实验表明，该方法在保持竞争力识别精度的同时，显著减少了感知步骤和运动成本，优于基于模型的基准方法。
原文摘要 · Abstract (English)
This paper presents a hybrid model-based and model-free framework for active multi-view target recognition using forward-looking sonar. A convolutional neural network (CNN) provides data-driven observation likelihoods, while Radon-based orientation estimation enables viewpoint-aware sensing without requiring angle annotations. During training, an information-gain-based reward guides a Proximal Policy Optimization (PPO) agent to learn a belief-aware viewpoint selection policy offline. At deployment, the learned policy performs real-time viewpoint selection using only CNN-based belief updates, eliminating the need for computationally expensive online POMDP tree search. Experiments on a marine-debris forward-looking sonar dataset demonstrate that the proposed approach achieves competitive recognition accuracy while reducing sensing steps and motion cost compared to model-based baselines.
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。