Abstract:Teleoperation is increasingly employed in environments where direct human access is difficult, such as hazardous exploration or surgical field. However, if the motion scale factor(MSF) intended to compensate for workspace-size differences is set inappropriately, repeated clutching operations and reduced precision can significantly raise cognitive load. This paper presents a shared controller that dynamically applies the MSF based on the user's intended motion scale. Inspired by human motor skills, the leader arm trajectory is divided into coarse(fast, large-range movements) and fine(precise, small-range movements), with three features extracted to train a fuzzy C-means(FCM) clustering model that probabilistically classifies the user's motion scale. Scaling the robot's motion accordingly reduces unnecessary repetition for large-scale movements and enables more precise control for fine operations. Incorporating recent trajectory data into model updates and offering user feedback for adjusting the MSF range and response speed allows mutual adaptation between user and system. In peg transfer experiments, compared to using a fixed single scale, the proposed approach demonstrated improved task efficiency(number of clutching and task completion time decreased 38.46% and 11.96% respectively), while NASA-TLX scores confirmed a meaningful reduction(58.01% decreased) in cognitive load. This outcome suggests that a user-intent-based motion scale adjustment can effectively enhance both efficiency and precision in teleoperation.
Abstract:Virtual Try-On (VTON) technology allows users to visualize how clothes would look on them without physically trying them on, gaining traction with the rise of digitalization and online shopping. Traditional VTON methods, often using Generative Adversarial Networks (GANs) and Diffusion models, face challenges in achieving high realism and handling dynamic poses. This paper introduces Outfitting Diffusion with Pose Guided Condition (ODPG), a novel approach that leverages a latent diffusion model with multiple conditioning inputs during the denoising process. By transforming garment, pose, and appearance images into latent features and integrating these features in a UNet-based denoising model, ODPG achieves non-explicit synthesis of garments on dynamically posed human images. Our experiments on the FashionTryOn and a subset of the DeepFashion dataset demonstrate that ODPG generates realistic VTON images with fine-grained texture details across various poses, utilizing an end-to-end architecture without the need for explicit garment warping processes. Future work will focus on generating VTON outputs in video format and on applying our attention mechanism, as detailed in the Method section, to other domains with limited data.
Abstract:In the physical sciences, there is an increased need for robust feature representations of image data: image acquisition, in the generalized sense of two-dimensional data, is now widespread across a large number of fields, including quantum information science, which we consider here. While traditional image features are widely utilized in such cases, their use is rapidly being supplanted by Neural Network-based techniques that often sacrifice explainability in exchange for high accuracy. To ameliorate this trade-off, we propose a synthetic data-based technique that results in explainable features. We show, using Explainable Boosting Machines (EBMs), that this method offers superior explainability without sacrificing accuracy. Specifically, we show that there is a meaningful benefit to this technique in the context of quantum dot tuning, where human intervention is necessary at the current stage of development.