Florian T. Pokorny

Dex-Net 1.0: A cloud-based network of 3D objects for robust grasp planning using a Multi-Armed Bandit model with correlated rewards

This paper presents the Dexterity Network (Dex-Net) 1.0, a dataset of 3D object models and a sampling-based planning algorithm to explore how Cloud Robotics can be used for robust grasp planning. The algorithm uses a Multi- Armed Bandit model with correlated rewards to leverage prior grasps and 3D object models in a growing dataset that currently includes over 10,000 unique 3D object models and 2.5 million parallel-jaw grasps. Each grasp includes an estimate of the probability of force closure under uncertainty in object and gripper pose and friction. Dex-Net 1.0 uses Multi-View Convolutional Neural Networks (MV-CNNs), a new deep learning method for 3D object classification, to provide a similarity metric between objects, and the Google Cloud Platform to simultaneously run up to 1,500 virtual cores, reducing experiment runtime by up to three orders of magnitude. Experiments suggest that correlated bandit techniques can use a cloud-based network of object models to significantly reduce the number of samples required for robust grasp planning. We report on system sensitivity to variations in similarity metrics and in uncertainty in pose and friction. Code and updated information is available at http://berkeleyautomation.github.io/dex-net/.

Classical grasp quality evaluation: New algorithms and theory

This paper investigates theoretical properties of a well-known L¹ grasp quality measure Q whose approximation Q^-_l is commonly used for the evaluation of grasps and where the precision of Q^-_l depends on an approximation of a cone by a convex polyhedral cone with l edges. We prove the Lipschitz continuity of Q and provide an explicit Lipschitz bound that can be used to infer the stability of grasps lying in a neighbourhood of a known grasp. We think of Q^-_l as a lower bound estimate to Q and describe an algorithm for computing an upper bound Q⁺. We provide worst-case error bounds relating Q and Q^-_l. Furthermore, we develop a novel grasp hypothesis rejection algorithm which can exclude unstable grasps much faster than current implementations. Our algorithm is based on a formulation of the grasp quality evaluation problem as an optimization problem, and we show how our algorithm can be used to improve the efficiency of sampling based grasp hypotheses generation methods.

Grasp Moduli Spaces

We present a new approach for modelling grasping using an integrated space of grasps and shapes. In particular, we introduce an infinite dimensional space, the Grasp Moduli Space, which represents shapes and grasps in a continuous manner. We define a metric on this space allowing us to formalize ‘nearby’ grasp/shape configurations and we discuss continuous deformations of such configurations. We work in particular with surfaces with cylindrical coordinates and analyse the stability of a popular L1 grasp quality measure Ql under continuous deformations of shapes and grasps. We experimentally determine bounds on the maximal change of Ql in a small neighbourhood around stable grasps with grasp quality above a threshold. In the case of surfaces of revolution, we determine stable grasps which correspond to grasps used by humans and develop an efficient algorithm for generating those grasps in the case of three contact points. We show that sufficiently stable grasps stay stable under small deformations. For larger deformations, we develop a gradient-based method that can transfer stable grasps between different surfaces. Additionally, we show in experiments that our gradient method can be used to find stable grasps on arbitrary surfaces with cylindrical coordinates by deforming such surfaces towards a corresponding ‘canonical’ surface of revolution.

SHIV: Reducing supervisor burden in DAgger using support vectors for efficient learning from demonstrations in high dimensional state spaces

Online learning from demonstration algorithms such as DAgger can learn policies for problems where the system dynamics and the cost function are unknown. However they impose a burden on supervisors to respond to queries each time the robot encounters new states while executing its current best policy. The MMD-IL algorithm reduces supervisor burden by filtering queries with insufficient discrepancy in distribution and maintaining multiple policies. We introduce the SHIV algorithm (Svm-based reduction in Human InterVention), which converges to a single policy and reduces supervisor burden in non-stationary high dimensional state distributions. To facilitate scaling and outlier rejection, filtering is based on a measure of risk defined in terms of distance to an approximate level set boundary defined by a One Class support vector machine. We report on experiments in three contexts: 1) a driving simulator with a 27,936 dimensional visual feature space, 2) a push-grasping in clutter simulation with a 22 dimensional state space, and 3) physical surgical needle insertion with a 16 dimensional state space. Results suggest that SHIV can efficiently learn policies with up to 70% fewer queries that DAgger.

Cohomological learning of periodic motion

This work develops a novel framework which can automatically detect, parameterize and interpolate periodic motion patterns obtained from a motion capture sequence. Using our framework, periodic motions such as walking and running gaits or any motion sequence with periodic structure such as cleaning, dancing etc. can be detected automatically and without manual marking of the period start and end points. Our approach constructs an intrinsic parameterization of the motion and is computationally fast. Using this parameterization, we are able generate prototypical periodic motions. Additionally, we are able to interpolate between various motions, yielding a rich class of ‘mixed’ periodic actions. Our approach is based on ideas from applied algebraic topology. In particular, we apply a novel persistent cohomology based method for the first time in a graphics application which enables us to recover circular coordinates of motions. We also develop a suitable notion of homotopy which can be used to interpolate between periodic motion patterns. Our framework is directly applicable to the construction of walk cycles for animating character motions with motion graphs or state machine driven animation engines and processed our examples at an average speed of 11.78 frames per secondGraphical abstract

Robot grasping in clutter: Using a hierarchy of supervisors for learning from demonstrations

For applications such as Amazon warehouse order fulfillment, robots must grasp a desired object amid clutter: other objects that block direct access. This can be difficult to program explicitly due to uncertainty in friction and push mechanics and the variety of objects that can be encountered. Deep Learning networks combined with Online Learning from Demonstration (LfD) algorithms such as DAgger and SHIV have potential to learn robot control policies for such tasks where the input is a camera image and system dynamics and the cost function are unknown. To explore this idea, we introduce a version of the grasping in clutter problem where a yellow cylinder must be grasped by a planar robot arm amid extruded objects in a variety of shapes and positions. To reduce the burden on human experts to provide demonstrations, we propose using a hierarchy of three levels of supervisors: a fast motion planner that ignores obstacles, crowd-sourced human workers who provide appropriate robot control values remotely via online videos, and a local human expert. Physical experiments suggest that with 160 expert demonstrations, using the hierarchy of supervisors can increase the probability of a successful grasp (reliability) from 55% to 90%.

Multiscale Topological Trajectory Classification with Persistent Homology

Topological approaches to studying equivalence classes of trajectories in a configuration space have recently received attention in robotics since they allow a robot to reason about trajectories at a high level of abstraction. While recent work has approached the problem of topological motion planning under the assumption that the configuration space and obstacles within it are explicitly described in a noise-free manner, we focus on trajectory classification and present a sampling-based approach which can handle noise, which is applicable to general configuration spaces and which relies only on the availability of collision free samples. Unlike previous sampling-based approaches in robotics which use graphs to capture information about the path-connectedness of a configuration space, we construct a multiscale approximation of neighborhoods of the collision free configurations based on filtrations of simplicial complexes. Our approach thereby extracts additional homological information which is essential for a topological trajectory classification. By computing a basis for the first persistent homology groups, we obtain a multiscale classification algorithm for trajectories in configuration spaces of arbitrary dimension. We furthermore show how an augmented filtration of simplicial complexes based on a cost function can be defined to incorporate additional constraints. We present an evaluation of our approach in 2, 3, 4 and 6 dimensional configuration spaces in simulation and using a Baxter robot.

Topological trajectory classification with filtrations of simplicial complexes and persistent homology

In this work, we present a sampling-based approach to trajectory classification which enables automated high-level reasoning about topological classes of trajectories. Our approach is applicable to general configuration spaces and relies only on the availability of collision free samples. Unlike previous sampling-based approaches in robotics which use graphs to capture information about the path-connectedness of a configuration space, we construct a multiscale approximation of neighborhoods of the collision free configurations based on filtrations of simplicial complexes. Our approach thereby extracts additional homological information which is essential for a topological trajectory classification. We propose a multiscale classification algorithm for trajectories in configuration spaces of arbitrary dimension and for sets of trajectories starting and ending in two fixed points. Using a cone construction, we then generalize this approach to classify sets of trajectories even when trajectory start and end points are allowed to vary in path-connected subsets. We furthermore show how an augmented filtration of simplicial complexes based on an arbitrary function on the configuration space, such as a costmap, can be defined to incorporate additional constraints. We present an evaluation of our approach in 2-, 3-, 4- and 6-dimensional configuration spaces in simulation and in real-world experiments using a Baxter robot and motion capture data.

Combinatorial optimization for hierarchical contact-level grasping

We address the problem of generating force-closed point contact grasps on complex surfaces and model it as a combinatorial optimization problem. Using a multilevel refinement metaheuristic, we maximize the quality of a grasp subject to a reachability constraint by recursively forming a hierarchy of increasingly coarser optimization problems. A grasp is initialized at the top of the hierarchy and then locally refined until convergence at each level. Our approach efficiently addresses the high dimensional problem of synthesizing stable point contact grasps while resulting in stable grasps from arbitrary initial configurations. Compared to a sampling-based approach, our method yields grasps with higher grasp quality. Empirical results are presented for a set of different objects. We investigate the number of levels in the hierarchy, the computational complexity, and the performance relative to a random sampling baseline approach.

Supervised Hierarchical Dirichlet Processes with Variational Inference

We present an extension to the Hierarchical Dirichlet Process (HDP), which allows for the inclusion of supervision. Our model marries the non-parametric benefits of HDP with those of Supervised Latent Dirichlet Allocation (SLDA) to enable learning the topic space directly from data while simultaneously including the labels within the model. The proposed model is learned using variational inference which allows for the efficient use of a large training dataset. We also present the online version of variational inference, which makes the method scalable to very large datasets. We show results comparing our model to a traditional supervised parametric topic model, SLDA, and show that it outperforms SLDA on a number of benchmark datasets.