Imitation learning (IL) algorithms typically distil experience into parametric behavior policies to mimic expert demonstrations. With limited experience previous methods often struggle and cannot accurately align the current state with expert demonstrations, particularly in tasks that are characterised by partial observations or dynamic object deformations. We consider imitation learning in deformable mobile manipulation with an ego-centric limited field of view and introduce a novel IL approach called DeMoBot that directly retrieves observations from demonstrations. DeMoBot utilizes vision foundation models to identify relevant expert data based on visual similarity and matches the current trajectory with demonstrated trajectories using trajectory similarity and forward reachability constraints to select suitable sub-goals. A goal-conditioned motion generation policy shall guide the robot to the sub-goal until the task is completed. We evaluate DeMoBot using a Spot robot in several simulated and real-world settings, demonstrating its effectiveness and generalizability. DeMoBot outperforms baselines with only 20 demonstrations, attaining high success rates in gap covering (85% simulation, 80% real-world) and table uncovering (87.5% simulation, 70% real-world), while showing promise in complex tasks like curtain opening (47.5% simulation, 35% real-world). Additional details are available at: https://sites.google.com/view/demobot-fewshot/home

翻译：模仿学习（IL）算法通常将经验提炼为参数化行为策略以模仿专家演示。在经验有限的情况下，先前的方法往往难以准确地将当前状态与专家演示对齐，尤其是在以部分观测或动态物体变形为特征的任务中。我们研究了在具有以自我为中心且视野受限的可变形移动操作中的模仿学习，并提出了一种名为DeMoBot的新型IL方法，该方法直接从演示中检索观测。DeMoBot利用视觉基础模型基于视觉相似性识别相关的专家数据，并通过轨迹相似性和前向可达性约束将当前轨迹与演示轨迹进行匹配，以选择合适的子目标。一个以目标为条件的运动生成策略将引导机器人到达子目标，直至任务完成。我们在多个模拟和真实世界场景中使用Spot机器人评估了DeMoBot，证明了其有效性和泛化能力。DeMoBot仅需20次演示即超越基线方法，在缝隙覆盖（模拟85%，真实世界80%）和桌面覆盖物移除（模拟87.5%，真实世界70%）任务中取得了高成功率，同时在复杂任务如窗帘拉开（模拟47.5%，真实世界35%）中也展现出潜力。更多细节请访问：https://sites.google.com/view/demobot-fewshot/home