| 1. |
- Mitsioni, Ioanna, 1991-, et al.
(författare)
-
Data-Driven Model Predictive Control for the Contact-Rich Task of Food Cutting
- 2019
-
Ingår i: IEEE-RAS International Conference on Humanoid Robots. - : IEEE Computer Society. - 9781538676301 ; 2019-October, s. 244-250
-
Konferensbidrag (refereegranskat)abstract
- Modelling of contact-rich tasks is challenging and cannot be entirely solved using classical control approaches due to the difficulty of constructing an analytic description of the contact dynamics. Additionally, in a manipulation task like food-cutting, purely learning-based methods such as Reinforcement Learning, require either a vast amount of data that is expensive to collect on a real robot, or a highly realistic simulation environment, which is currently not available. This paper presents a data-driven control approach that employs a recurrent neural network to model the dynamics for a Model Predictive Controller. We build upon earlier work limited to torque-controlled robots and redefine it for velocity controlled ones. We incorporate force/torque sensor measurements, reformulate and further extend the control problem formulation. We evaluate the performance on objects used for training, as well as on unknown objects, by means of the cutting rates achieved and demonstrate that the method can efficiently treat different cases with only one dynamic model. Finally we investigate the behavior of the system during force-critical instances of cutting and illustrate its adaptive behavior in difficult cases.
|
|
| 2. |
- Mitsioni, Ioanna, 1991-, et al.
(författare)
-
Interpretability in Contact-Rich Manipulation via Kinodynamic Images
- 2021
-
Ingår i: Proceedings - IEEE International Conference on Robotics and Automation. - : Institute of Electrical and Electronics Engineers (IEEE). - 1050-4729. ; 2021-May, s. 10175-10181
-
Konferensbidrag (refereegranskat)abstract
- Deep Neural Networks (NNs) have been widely utilized in contact-rich manipulation tasks to model the complicated contact dynamics. However, NN-based models are often difficult to decipher which can lead to seemingly inexplicable behaviors and unidentifiable failure cases. In this work, we address the interpretability of NN-based models by introducing the kinodynamic images. We propose a methodology that creates images from kinematic and dynamic data of contact-rich manipulation tasks. By using images as the state representation, we enable the application of interpretability modules that were previously limited to vision-based tasks. We use this representation to train a Convolutional Neural Network (CNN) and we extract interpretations with Grad-CAM to produce visual explanations. Our method is versatile and can be applied to any classification problem in manipulation tasks to visually interpret which parts of the input drive the model's decisions and distinguish its failure modes, regardless of the features used. Our experiments demonstrate that our method enables detailed visual inspections of sequences in a task, and high-level evaluations of a model's behavior.
|
|
| 3. |
- Antonova, Rika, et al.
(författare)
-
Global Search with Bernoulli Alternation Kernel for Task-oriented Grasping Informed by Simulation
- 2018
-
Ingår i: Proceedings of Machine Learning Research. - : PMLR. ; , s. 641-650, s. 641-650
-
Konferensbidrag (refereegranskat)abstract
- We develop an approach that benefits from large simulated datasets and takes full advantage of the limited online data that is most relevant. We propose a variant of Bayesian optimization that alternates between using informed and uninformed kernels. With this Bernoulli Alternation Kernel we ensure that discrepancies between simulation and reality do not hinder adapting robot control policies online. The proposed approach is applied to a challenging real-world problem of task-oriented grasping with novel objects. Our further contribution is a neural network architecture and training pipeline that use experience from grasping objects in simulation to learn grasp stability scores. We learn task scores from a labeled dataset with a convolutional network, which is used to construct an informed kernel for our variant of Bayesian optimization. Experiments on an ABB Yumi robot with real sensor data demonstrate success of our approach, despite the challenge of fulfilling task requirements and high uncertainty over physical properties of objects.
|
|
| 4. |
- Kokic, Mia, et al.
(författare)
-
Affordance detection for task-specific grasping using deep learning
- 2017
-
Ingår i: 2017 IEEE-RAS 17th International Conference on Humanoid Robotics (Humanoids). - : IEEE conference proceedings. - 9781538646786 ; , s. 91-98
-
Konferensbidrag (refereegranskat)abstract
- In this paper we utilize the notion of affordances to model relations between task, object and a grasp to address the problem of task-specific robotic grasping. We use convolutional neural networks for encoding and detecting object affordances, class and orientation, which we utilize to formulate grasp constraints. Our approach applies to previously unseen objects from a fixed set of classes and facilitates reasoning about which tasks an object affords and how to grasp it for that task. We evaluate affordance detection on full-view and partial-view synthetic data and compute task-specific grasps for objects that belong to ten different classes and afford five different tasks. We demonstrate the feasibility of our approach by employing an optimization-based grasp planner to compute task-specific grasps.
|
|
| 5. |
- Bekiroglu, Yasemin, 1982-, et al.
(författare)
-
Learning Tactile Characterizations Of Object- And Pose-specific Grasps
- 2011
-
Konferensbidrag (refereegranskat)abstract
- Our aim is to predict the stability of a grasp from the perceptions available to a robot before attempting to lift up and transport an object. The percepts we consider consist of the tactile imprints and the object-gripper configuration read before and until the robot’s manipulator is fully closed around an object. Our robot is equipped with multiple tactile sensing arrays and it is able to track the pose of an object during the application of a grasp. We present a kernel-logistic-regression model of pose- and touch-conditional grasp success probability which we train on grasp data collected by letting the robot experience the effect on tactile and visual signals of grasps suggested by a teacher, and letting the robot verify which grasps can be used to rigidly control the object. We consider models defined on several subspaces of our input data – e.g., using tactile perceptions or pose information only. Our experiment demonstrates that joint tactile and pose-based perceptions carry valuable grasp-related information, as models trained on both hand poses and tactile parameters perform better than the models trained exclusively on one perceptual input.
|
|
| 6. |
- Hang, Kaiyu, 1987-, et al.
(författare)
-
Hierarchical Fingertip Space : A Unified Framework for Grasp Planning and In-Hand Grasp Adaptation
- 2016
-
Ingår i: IEEE Transactions on robotics. - : IEEE Press. - 1552-3098 .- 1941-0468. ; 32:4, s. 960-972
-
Tidskriftsartikel (refereegranskat)abstract
- We present a unified framework for grasp planning and in-hand grasp adaptation using visual, tactile and proprioceptive feedback. The main objective of the proposed framework is to enable fingertip grasping by addressing problems of changed weight of the object, slippage and external disturbances. For this purpose, we introduce the Hierarchical Fingertip Space (HFTS) as a representation enabling optimization for both efficient grasp synthesis and online finger gaiting. Grasp synthesis is followed by a grasp adaptation step that consists of both grasp force adaptation through impedance control and regrasping/finger gaiting when the former is not sufficient. Experimental evaluation is conducted on an Allegro hand mounted on a Kuka LWR arm.
|
|
| 7. |
- Pokorny, Florian, et al.
(författare)
-
Grasp Moduli Spaces, Gaussian Processes and Multimodal Sensor Data
- 2014
-
Konferensbidrag (övrigt vetenskapligt/konstnärligt)abstract
- We discuss our recent and ongoing work on deformation-based grasp synthesis and transfer which formulates a joint space of shapes and grasps – a Grasp Moduli Space– within which both grasp configurations and object shapes can be continuously deformed. In this context, we propose the use of Gaussian Process-based implicit surface representations. These shape representations complement our previous work on using spherical harmonics as well as explicit cylindrical coordinates to model the object shape component of a Grasp Moduli Space. We provide a preliminary experiment with these Gaussian Processes showing how shape representations can be obtained using multimodal visual and haptic information and discuss how these representations can be continuously deformed for the purpose of transferring and generalizing known grasps.
|
|
| 8. |
- Stork, Johanes Andreas, et al.
(författare)
-
Learning Predictive State Representation for in-hand manipulation
- 2015
-
Ingår i: Proceedings - IEEE International Conference on Robotics and Automation. - : IEEE conference proceedings. - 1050-4729. ; , s. 3207-3214
-
Konferensbidrag (refereegranskat)abstract
- We study the use of Predictive State Representation (PSR) for modeling of an in-hand manipulation task through interaction with the environment. We extend the original PSR model to a new domain of in-hand manipulation and address the problem of partial observability by introducing new kernel-based features that integrate both actions and observations. The model is learned directly from haptic data and is used to plan series of actions that rotate the object in the hand to a specific configuration by pushing it against a table. Further, we analyze the model's belief states using additional visual data and enable planning of action sequences when the observations are ambiguous. We show that the learned representation is geometrically meaningful by embedding labeled action-observation traces. Suitability for planning is demonstrated by a post-grasp manipulation example that changes the object state to multiple specified target configurations.
|
|
| 9. |
- Hang, Kaiyu, et al.
(författare)
-
Hierarchical Fingertip Space for Synthesizing Adaptable Fingertip Grasps
- 2014
-
Konferensbidrag (övrigt vetenskapligt/konstnärligt)abstract
- The ability to synthesize and execute fingertip grasps are bases for dexterous in-hand manipulation. Reliable fingertip grasping is difficult to achieve due to noise and uncertainties in object and hand model, as well as hand control etc. Moreover, in many cases it is desirable to employ an adaptive approach that can deal with changed external forces. In this paper, we propose an approach to jointly optimize stability, adaptability, and reachability of grasps using combinatorial optimization for a hierarchical representation of the fingertip space. To illustrate our approach, we demonstrate an example synthesized by the proposed framework and executed by an Allegro hand. We also show how it is adapted when a perturbation is applied.
|
|
| 10. |
- Bekiroglu, Yasemin, 1982, et al.
(författare)
-
Joint Observation of Object Pose and Tactile Imprints for Online Grasp Stability Assessment
- 2011
-
Konferensbidrag (övrigt vetenskapligt/konstnärligt)abstract
- This paper studies the viability of concurrent object pose tracking and tactile sensing for assessing grasp stability on a physical robotic platform. We present a kernel logistic regression model of pose- and touch-conditional grasp success probability. Models are trained on grasp data which consist of (1) the pose of the gripper relative to the object, (2) a tactile description of the contacts between the object and the fully-closed gripper, and (3) a binary description of grasp feasibility, which indicates whether the grasp can be used to rigidly control the object. The data is collected by executing grasps demonstrated by a human on a robotic platform composed of an industrial arm, a three-finger gripper equipped with tactile sensing arrays, and a vision-based object pose tracking system. The robot is able to track the pose of an object while it is grasping it, and it can acquire grasp tactile imprints via pressure sensor arrays mounted on its gripper’s fingers. We consider models defined on several subspaces of our input data – using tactile perceptions or gripper poses only. Models are optimized and evaluated with f-fold cross-validation. Our preliminary results show that stability assessments based on both tactile and pose data can provide better rates than assessments based on tactile data alone.
|
|
