Abstract:This work introduces a motion retargeting approach for legged robots, which aims to create motion controllers that imitate the fine behavior of animals. Our approach, namely spatio-temporal motion retargeting (STMR), guides imitation learning procedures by transferring motion from source to target, effectively bridging the morphological disparities by ensuring the feasibility of imitation on the target system. Our STMR method comprises two components: spatial motion retargeting (SMR) and temporal motion retargeting (TMR). On the one hand, SMR tackles motion retargeting at the kinematic level by generating kinematically feasible whole-body motions from keypoint trajectories. On the other hand, TMR aims to retarget motion at the dynamic level by optimizing motion in the temporal domain. We showcase the effectiveness of our method in facilitating Imitation Learning (IL) for complex animal movements through a series of simulation and hardware experiments. In these experiments, our STMR method successfully tailored complex animal motions from various media, including video captured by a hand-held camera, to fit the morphology and physical properties of the target robots. This enabled RL policy training for precise motion tracking, while baseline methods struggled with highly dynamic motion involving flying phases. Moreover, we validated that the control policy can successfully imitate six different motions in two quadruped robots with different dimensions and physical properties in real-world settings.
Abstract:With the development of the Internet and the accumulation of information on the web, users use a search engine to easily obtain the desired information. A query suggestion is one of the main services provided by a search engine, and is very important for improving search performance, creating efficient queries, and reducing search time. However, there are search engines that do not support the query suggestion service. Under such engines, if users want to perform a search, they would have much difficulties in effectively performing the search. In this paper, to tackle the problem, we propose and develop a metasuggestion engine that crawls suggested search queries from search engines with a suggestion service, applies a re-ranking algorithm, and provides the suggested search queries in the form of an extension program on a web browser. Meta-suggestion engine are useful for users searching in engines that do not provide query suggestions, as they provide query suggestions wherever the user searches. We evaluate engines with relevance-based and predictive hit-based evaluation methods, showing that MSE produces good quality suggestions. We study improvements in target engine selection and re-ranking algorithms in future studies.