Frontiers in Robotics and AI最新文献

Dance plays a vital role in human societies across time and culture, with different communities having invented different systems for artistic expression through movement (genres). Differences between genres can be described by experts in words and movements, but these descriptions can only be appreciated by people with certain background abilities. Existing dance notation schemes could be applied to describe genre-differences, however they fall substantially short of being able to capture the important details of movement across a wide spectrum of genres. Our knowledge and practice around dance would benefit from a general, quantitative and human-understandable method of characterizing meaningful differences between aspects of any dance style; a computational kinematics of dance. Here we introduce and apply a novel system for encoding bodily movement as 17 macroscopic, interpretable features, such as expandedness of the body or the frequency of sharp movements. We use this encoding to analyze Hip Hop Dance genres, in part by building a low-cost machine-learning classifier that distinguishes genre with high accuracy. Our study relies on an open dataset (AIST++) of pose-sequences from dancers instructed to perform one of ten Hip Hop genres, such as Breakdance, Popping, or Krump. For comparison we evaluate moderately experienced human observers at discerning these sequence's genres from movements alone (38% where chance = 10%). The performance of a baseline, Ridge classifier model was fair (48%) and that of the model resulting from our automated machine learning pipeline was strong (76%). This indicates that the selected features represent important dimensions of movement for the expression of the attitudes, stories, and aesthetic values manifested in these dance forms. Our study offers a new window into significant relations of similarity and difference between the genres studied. Given the rich, complex, and culturally shaped nature of these genres, the interpretability of our features, and the lightweight techniques used, our approach has significant potential for generalization to other movement domains and movement-related applications.

Reminding is often identified as a central function of socially assistive robots in the healthcare sector. The robotic reminders are supposed to help people with memory impairments to remember to take their medicine, to drink and eat, or to attend appointments. Such standalone reminding technologies can, however, be too demanding for people with memory injuries. In a co-creation process, we developed an individual reminder robot together with a person with traumatic brain injury and her care personnel. During this process, we learned that while current research describe reminding as a prototypical task for socially assistive robots, there is no clear definition of what constitutes a reminder nor that it is based on complex sequences of interactions that evolve over time and space, across different actions, actors and technologies. Based on our data from the co-creation process and the first deployment, we argue for a shift towards a sequential and socially distributed character of reminding. Understanding socially assistive robots as rehabilitative tools for people with memory impairment, they need to be reconsidered as interconnected elements in institutional care practices instead of isolated events for the remindee.

In robotics, active exploration and learning in uncertain environments must take into account safety, as the robot may otherwise damage itself or its surroundings. This paper presents a method for safe active search using Bayesian optimization and control barrier functions. As robot paths undertaken during sampling are continuous, we consider an informative continuous expected improvement acquisition function. To safely bound the contact forces between the robot and its surroundings, we leverage exponential control barrier functions, utilizing the derivative of the force in the contact model to increase robustness to uncertainty in the contact boundary. Our approach is demonstrated on a fully autonomous robot for ultrasound scanning of rheumatoid arthritis (RA). Here, active search is a critical component of ensuring high image quality. Furthermore, bounded contact forces between the ultrasound probe and the patient ensure patient safety and better scan quality. To the best of our knowledge, our results are both the first demonstration of safe active search on a fully autonomous robot for ultrasound scanning of rheumatoid arthritis and the first experimental evaluation of bounding contact forces in the context of medical robotics using control barrier functions. The results show that when search time is limited to less than 60 s, informative continuous expected improvement leads to a 92% success, a 13% improvement compared to expected improvement. Meanwhile, exponential control barrier functions can limit the force applied by the robot to under 5 N, even in cases where the contact boundary is specified incorrectly by -1 or +4 mm.

This paper focuses on the design of Convolution Neural Networks to visually guide an autonomous Unmanned Aerial Vehicle required to inspect power towers. The network is required to precisely segment images taken by a camera mounted on a UAV in order to allow a motion module to generate collision-free and inspection-relevant manoeuvres of the UAV along different types of towers. The images segmentation process is particularly challenging not only because of the different structures of the towers but also because of the enormous variability of the background, which can vary from the uniform blue of the sky to the multi-colour complexity of a rural, forest, or urban area. To be able to train networks that are robust enough to deal with the task variability, without incurring into a labour-intensive and costly annotation process of physical-world images, we have carried out a comparative study in which we evaluate the performances of networks trained either with synthetic images (i.e., the synthetic dataset), physical-world images (i.e., the physical-world dataset), or a combination of these two types of images (i.e., the hybrid dataset). The network used is an attention-based U-NET. The synthetic images are created using photogrammetry, to accurately model power towers, and simulated environments modelling a UAV during inspection of different power towers in different settings. Our findings reveal that the network trained on the hybrid dataset outperforms the networks trained with the synthetic and the physical-world image datasets. Most notably, the networks trained with the hybrid dataset demonstrates a superior performance on multiples evaluation metrics related to the image-segmentation task. This suggests that, the combination of synthetic and physical-world images represents the best trade-off to minimise the costs related to capturing and annotating physical-world images, and to maximise the task performances. Moreover, the results of our study demonstrate the potential of photogrammetry in creating effective training datasets to design networks to automate the precise movement of visually-guided UAVs.