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.

翻译：本文提出了一种基于混合模型与无模型的主动多视角目标识别框架，用于前视声纳系统。卷积神经网络（CNN）提供数据驱动的观测似然估计，而基于Radon变换的方向估计方法无需角度标注即可实现视点感知。训练阶段，基于信息增益的奖励机制引导近端策略优化（PPO）智能体离线学习具有信念感知的视点选择策略。部署时，所学策略仅依赖CNN信念更新执行实时视点选择，无需计算昂贵的在线部分可观测马尔可夫决策过程（POMDP）树搜索。在海洋垃圾前视声纳数据集上的实验表明，与基于模型的基线方法相比，所提方法在保持竞争性识别精度的同时，有效减少了感知步数与运动成本。