17 accepted papers
We introduce Euler–Poincaré Neural Dynamics (EPND), a geometric-mechanics–driven framework that redefines how Koopman-type neural models learn dynamical systems. Unlike conventional operator-learning methods that rely on function-space linearization, EPND places geometric mechanics at the core of it…
We present a Lie-algebraic approach to model Koopman dynamics that integrates algebraic structure with computational scalability. The proposed formulation constrains the neural generators to evolve within prescribed Lie subalgebras and constructs finite-time flows through a neural Magnus expansion c…
We propose a novel class of neural differential equation models called mean-field continuous sequence predictors (MFPs) for efficiently generating continuous sequences with potentially infinite-order complexity. To address complex inductive biases in time-series data, we employ mean-field dynamics s…
This study addresses the challenge of implementing proprioceptive and kinesthetic (PK) feedback in robotic hands, essential for grasping and manipulation tasks in unstructured environments. We developed a compact modular actuator featuring a low-module, high-transmission-ratio multistage gear mechan…
We study the linear bandit problem that accounts for partially observable features. Without proper handling, unobserved features can lead to linear regret in the decision horizon $T$, as their influence on rewards is unknown. To tackle this challenge, we propose a novel theoretical framework and an…
This study addresses the challenge of implementing proprioceptive and kinesthetic (PK) feedback in robotic hands, essential for grasping and manipulation tasks in unstructured environments. We developed a compact modular actuator featuring a low-module, high-transmission-ratio multistage gear mechan
We introduce Neural Hamiltonian Diffusion, a unified framework for learning stochastic Hamiltonian dynamics on differentiable manifolds. While Hamiltonian Neural Networks (HNNs) model conservative systems in flat Euclidean space, they fail to account for geometric structure and intrinsic stochastici…
Recent advances in generative diffusion models have enabled text-controlled synthesis of realistic and diverse images with impressive quality. Despite these remarkable advances, the application of text-to-image generative models in computer vision for standard visual recognition tasks remains limite…
In this paper, we introduce a new class of score-based generative models (SGMs) designed to handle high-cardinality data distributions by leveraging concepts from mean-field theory. We present mean-field chaos diffusion models (MF-CDMs), which address the curse of dimensionality inherent in high-car…
For realistic talking head generation, creating natural head motion while maintaining accurate lip synchronization is essential. To fulfill this challenging task, we propose DisCoHead, a novel method to disentangle and control head pose and facial expressions without supervision. DisCoHead uses a si…
We develop a novel hall-effect-based six-axis force/torque (F/T) tactile sensor integrated into the fingertip of robotic hands. When the robotic hands performs the grasping tasks in an unstructured environment, the visual information plays a main role in sensing the external properties of the object…
Modeling spatiotemporal dynamics with neural differential equations has become a major line of research that opens new ways to handle various real-world scenarios (e.g., missing observations, irregular times, etc.). Despite such progress, most existing methods still face challenges in providing a ge…
We design a central controller system (CCS) and a tele-controlled system (TCS) with an aim of developing the integrated tele-monitoring/operation system that can enable the medical staff to tele-monitor the state of therapeutic devices utilized in the isolation intensive care unit (ICU) and to tele-
A variety of effective face-swap and face-reenactment methods have been publicized in recent years, democratizing the face synthesis technology to a great extent. Videos generated as such have come to be called deepfakes with a negative connotation, for various social problems they have caused. Faci…
We develop a tactile information-based pinch-grasp strategy enabling a robot hand to adaptively grasp easily deformable soft objects. When a robot hand has to perform grasping tasks, the grasp planner develops the grasping strategy based on visual information. However, the intrinsic properties of th
We develop a surgical forceps-driver and a haptic feedback master device that could be applied to a tele-operated microsurgical robotic system for peripheral nerve surgery. Considering that the neurological damage could occur to the peripheral nerve when an excessive force is applied to the nerve ti
Hybrid analog/digital precoding in the downlink of multiuser massive MIMO systems can reduce the number of RF chains hence reducing total cost and improving power efficiency. Having few RF chains, however, makes it difficult for a base station to acquire instantaneous channel state information acros…