24 accepted papers
In this letter, we propose a novel reinforcement learning framework that enables zero-shot policy adaptation in environments with unseen, dynamically changing obstacles. Adopting the idea that learning a policy capable of generating diverse actions is key to achieving such adaptability, our primary
The performance of learning-based object detection algorithms, which attempt to both classify and locate objects within images, is determined largely by the quality of the annotated dataset used for training. Two types of labelling noises are prevalent: objects that are incorrectly classified (categ…
Effective movement primitives should be capable of encoding and generating a rich repertoire of trajectories conditioned on task-defining parameters such as vision or language inputs. While recent methods based on the motion manifold hypothesis, which assumes that a set of trajectories lies on a low
Articulated object recognition -- the task of identifying both the geometry and kinematic joints of objects with movable parts -- is essential for enabling robots to interact with everyday objects such as doors and laptops. However, existing approaches often rely on strong assumptions, such as a kno…
Traditional methods for synthesizing 6-DoF grasp poses from 3D observations often rely on geometric heuristics, resulting in poor generalizability, limited grasp options, and higher failure rates. Recently, data-driven methods have been proposed that use generative models to learn the distribution o…
When using an autoencoder to learn the low-dimensional manifold of high-dimensional data, it is crucial to find the latent representations that preserve the geometry of the data manifold. However, most existing studies assume a Euclidean nature for the high-dimensional data space, which is arbitrary…
We present a maximum entropy inverse reinforcement learning (IRL) approach for improving the sample quality of diffusion generative models, especially when the number of generation time steps is small. Similar to how IRL trains a policy based on the reward function learned from expert demonstrations…
Recognizing and manipulating transparent tableware from partial view RGB image observations is made challenging by the difficulty in obtaining reliable depth measurements of transparent objects. In this paper we present the Transparent Tableware SuperQuadric Network (T$^2$SQNet), a neural network m…
We present a new method of training energy-based models (EBMs) for anomaly detection that leverages low-dimensional structures within data. The proposed algorithm, Manifold Projection-Diffusion Recovery (MPDR), first perturbs a data point along a low-dimensional manifold that approximates the traini…
Existing movement primitive models for the most part focus on representing and generating a single trajectory for a given task, limiting their adaptability to situations in which unforeseen obstacles or new constraints may arise. In this work we propose Motion Manifold Primitives (MMP), a movement p…
Regularization is almost {\it de rigueur} when designing autoencoders that are sparse and robust to noise. Given the recent surge of interest in machine learning problems involving non-Euclidean data, in this paper we address the regularization of autoencoders on curved spaces. We show that by ignor…
Finding and grasping a target object on a cluttered shelf, especially when the target is occluded by other unknown objects and initially invisible, remains a significant challenge in robotic manipulation. While there have been advances in finding the target object by rearranging surrounding objects…
Motion planning for robot manipulation systems operating in complex environments remains a challenging problem. It requires the evaluation of both the collision distance and its derivative. Owing to its computational complexity, recent studies have attempted to utilize data-driven approaches to lear…
Kernel density estimation (KDE) is integral to a range of generative and discriminative tasks in machine learning. Drawing upon tools from the multidimensional calculus of variations, we derive an optimal weight function that reduces bias in standard kernel density estimates for density ratios, lead…
It has been observed that the generalization performance of neural networks correlates with the sharpness of their loss landscape. Dinh et al. (2017) have observed that existing formulations of sharpness measures fail to be invariant with respect to scaling and reparametrization. While some scale-in…
With recent advances in rapid prototyping and mechatronics, the problem of simultaneous design and motion optimization, or the co-design problem, is becoming more and more relevant in robotics. For reasons of computational tractability, all existing methods use simplified approximations of a robot’s
The recent success of autoencoders for representation learning can be traced in large part to the addition of a regularization term. Such regularized autoencoders ``constrain" the representation so as to prevent overfitting to the data while producing a parsimonious generative model. A regularized a…
For tabletop object manipulation tasks, learning an accurate pushing dynamics model, which predicts the objects' motions when a robot pushes an object, is very important. In this work, we claim that an ideal pushing dynamics model should have the SE(2)-equivariance property, i.e., if tabletop object…
Robots operating in close proximity to humans require fast and reliable detection of collisions, which can range from sharp impacts (hard collisions) to pulling-pushing-catching motions (soft collisions). Because joint torque sensors can be costly, the external joint torques caused by collisions are
Vanilla autoencoders often produce manifolds that overfit to noisy training data, or have the wrong local connectivity and geometry. Autoencoder regularization techniques, e.g., the denoising autoencoder, have had some success in reducing overfitting, whereas recent graph-based methods that exploit…
A neural network for generating coordinated reach-grasp motions is proposed, based on a type of generative neural network called the conditional restricted Boltzmann machine (CRBM). Given demonstrations of humans reaching and grasping various target objects of different shapes and poses, a mixture-t
Notwithstanding the seemingly straightforward nature of the inertial parameter identification problem for multibody systems-the most common formulation is as a linear least-squares estimation problem-existing methods, especially for complex high-dof systems subject to nonpersistent and noisy measure
Existing adaptive robot control laws typically require an engineering choice of a constant adaptation gain matrix, which often involves repeated and time-consuming trial and error. Moreover, physical consistency of the estimated inertial parameters or the uniform positive definiteness of the estimat…
Motion planning algorithms that rely upon the randomly exploring random tree (RRT) typically require the user to choose an appropriate stepsize; this is generally a highly problem-dependent and time-consuming process requiring trial and error. We propose an adaptive stepsize RRT path planning algori