Learning Based Robot Vision

Industrial robots carry out simple tasks in customized environments for which it is typical that nearly all e?ector movements can be planned during an - line phase. A continual control based on sensory feedback is at most necessary at e?ector positions near target locations utilizing torque or haptic sensors. It is desirable to develop new-generation robots showing higher degrees of autonomy for solving high-level deliberate tasks in natural and dynamic en- ronments. Obviously, camera-equipped robot systems, which take and process images and make use of the visual data, can solve more sophisticated robotic tasks. The development of a (semi-) autonomous camera-equipped robot must be grounded on an infrastructure, based on which the system can acquire and/or adapt task-relevant competences autonomously. This infrastructure consists of technical equipment to support the presentation of real world training samples, various learning mechanisms for automatically acquiring function approximations, and testing methods for evaluating the quality of the learned functions. Accordingly, to develop autonomous camera-equipped robot systems one must ?rst demonstrate relevant objects, critical situations, and purposive situation-action pairs in an experimental phase prior to the application phase. Secondly, the learning mechanisms are responsible for - quiring image operators and mechanisms of visual feedback control based on supervised experiences in the task-relevant, real environment. This paradigm of learning-based development leads to the concepts of compatibilities and manifolds. Compatibilities are general constraints on the process of image formation which hold more or less under task-relevant or accidental variations of the imaging conditions.

Robotic systems consist of object or scene recognition, vision-based motion control, vision-based mapping, and dense range sensing, and are used for identification and navigation. As these computer vision and robotic connections continue to develop, the benefits of vision technology including savings, improved quality, reliability, safety, and productivity are revealed. Robotic Vision: Technologies for Machine Learning and Vision Applications is a comprehensive collection which highlights a solid framework for understanding existing work and planning future research. This book includes current research on the fields of robotics, machine vision, image processing and pattern recognition that is important to applying machine vision methods in the real world.

Industrial robots carry out simple tasks in customized environments for which it is typical that nearly all e?ector movements can be planned during an - line phase. A continual control based on sensory feedback is at most necessary at e?ector positions near target locations utilizing torque or haptic sensors. It is desirable to develop new-generation robots showing higher degrees of autonomy for solving high-level deliberate tasks in natural and dynamic en- ronments. Obviously, camera-equipped robot systems, which take and process images and make use of the visual data, can solve more sophisticated robotic tasks. The development of a (semi-) autonomous camera-equipped robot must be grounded on an infrastructure, based on which the system can acquire and/or adapt task-relevant competences autonomously. This infrastructure consists of technical equipment to support the presentation of real world training samples, various learning mechanisms for automatically acquiring function approximations, and testing methods for evaluating the quality of the learned functions. Accordingly, to develop autonomous camera-equipped robot systems one must ?rst demonstrate relevant objects, critical situations, and purposive situation-action pairs in an experimental phase prior to the application phase. Secondly, the learning mechanisms are responsible for - quiring image operators and mechanisms of visual feedback control based on supervised experiences in the task-relevant, real environment. This paradigm of learning-based development leads to the concepts of compatibilities and manifolds. Compatibilities are general constraints on the process of image formation which hold more or less under task-relevant or accidental variations of the imaging conditions.

Deep Learning for Robot Perception and Cognition introduces a broad range of topics and methods in deep learning for robot perception and cognition together with end-to-end methodologies. The book provides the conceptual and mathematical background needed for approaching a large number of robot perception and cognition tasks from an end-to-end learning point-of-view. The book is suitable for students, university and industry researchers and practitioners in Robotic Vision, Intelligent Control, Mechatronics, Deep Learning, Robotic Perception and Cognition tasks. Presents deep learning principles and methodologies Explains the principles of applying end-to-end learning in robotics applications Presents how to design and train deep learning models Shows how to apply deep learning in robot vision tasks such as object recognition, image classification, video analysis, and more Uses robotic simulation environments for training deep learning models Applies deep learning methods for different tasks ranging from planning and navigation to biosignal analysis

Assembled in this volume is a collection of some of the state-of-the-art methods that are using computer vision and machine learning techniques as applied in robotic applications. Currently there is a gap between research conducted in the computer vision and robotics communities. This volume discusses contrasting viewpoints of computer vision vs. robotics, and provides current and future challenges discussed from a research perspective.

Robot vision refers to the capability of a robot to visually perceive the environment and use this information for execution of various tasks. Visual feedback has been used extensively for robot navigation and obstacle avoidance. In the recent years, there are also examples that include interaction with people and manipulation of objects. In this paper, we review some of the work that goes beyond of using artificial landmarks and fiducial markers for the purpose of implementing visionbased control in robots. We discuss different application areas, both from the systems perspective and individual problems such as object tracking and recognition.

The field of robotic vision has advanced dramatically recently with the development of new range sensors. Tremendous progress has been made resulting in significant impact on areas such as robotic navigation, scene/environment understanding, and visual learning. This edited book provides a solid and diversified reference source for some of the most recent important advancements in the field of robotic vision. The book starts with articles that describe new techniques to understand scenes from 2D/3D data such as estimation of planar structures, recognition of multiple objects in the scene using different kinds of features as well as their spatial and semantic relationships, generation of 3D object models, approach to recognize partially occluded objects, etc. Novel techniques are introduced to improve 3D perception accuracy with other sensors such as a gyroscope, positioning accuracy with a visual servoing based alignment strategy for microassembly, and increasing object recognition reliability using related manipulation motion models. For autonomous robot navigation, different vision-based localization and tracking strategies and algorithms are discussed. New approaches using probabilistic analysis for robot navigation, online learning of vision-based robot control, and 3D motion estimation via intensity differences from a monocular camera are described. This collection will be beneficial to graduate students, researchers, and professionals working in the area of robotic vision.

Over the past five years robot vision has emerged as a subject area with its own identity. A text based on the proceedings of the Symposium on Computer Vision and Sensor-based Robots held at the General Motors Research Laboratories, Warren, Michigan in 1978, was published by Plenum Press in 1979. This book, edited by George G. Dodd and Lothar Rosso!, probably represented the first identifiable book covering some aspects of robot vision. The subject of robot vision and sensory controls (RoViSeC) occupied an entire international conference held in the Hilton Hotel in Stratford, England in May 1981. This was followed by a second RoViSeC held in Stuttgart, Germany in November 1982. The large attendance at the Stratford conference and the obvious interest in the subject of robot vision at international robot meetings, provides the stimulus for this current collection of papers. Users and researchers entering the field of robot vision for the first time will encounter a bewildering array of publications on all aspects of computer vision of which robot vision forms a part. It is the grey area dividing the different aspects of computer vision which is not easy to identify. Even those involved in research sometimes find difficulty in separating the essential differences between vision for automated inspection and vision for robot applications. Both of these are to some extent applications of pattern recognition with the underlying philosophy of each defining the techniques used.

In the past decade, researchers are looking at bringing robots into our daily lives, and automating services such as taxis, delivery, house-works, and even medical procedures. One of the major roadblocks in making this leap is the diversity and uncertainty in the environments that the robots need to work in. Machine perception, i.e. understanding of the environment through visual, audio, and contact signals, is indispensable in such diverse and uncertain environments, and is a hard problem in itself. Further, the environment is changing, due to human activities and other factors, and robots need to react to the changes quickly. Recent developments in deep learning, especially computer vision, has brought us closer to achieving the goal of bringing robots into our daily environments. However, deep learning methods require a large amount of data with annotated labels, and new datasets and annotations need to be collected for each new task. Deep reinforcement learning algorithms have also achieved good performance on a range of locomotion or manipulation tasks, but the amount of interactions required to train most algorithms is so large that it could take days even with parallel simulation engines. Highly data-efficient models and learning algorithms are needed to help robots learn faster and with less human effort. Additionally, when designing a learning-based solution to a robotics task, inference speed needs to be taken into consideration so that the robot can respond to changes quickly. This thesis introduces methods to improve training data efficiency and inference speed for vision-based robotic manipulation. To improve data efficiency of models, we analyze properties and structures of the specific problems, and build structural biases into the models based on the insights obtained. In addition, we demonstrate self-supervised learning of the perception model on real images, enabling robots to collect their own training data without requiring human annotations. To improve robots' response speed, when learning motion policies we design learning algorithms to always explicitly learn the distribution of promising actions, instead of learning an action evaluation function which requires online optimization during runtime. The proposed methods are integrated into end-to-end systems and tested on real robots on two tasks: vision-based robotic grasping, and rope manipulation and knotting.

The book is intended for advanced students in physics, mathematics, computer science, electrical engineering, robotics, engine engineering and for specialists in computer vision and robotics on the techniques for the development of vision-based robot projects. It focusses on autonomous and mobile service robots for indoor work, and teaches the techniques for the development of vision-based robot projects. A basic knowledge of informatics is assumed, but the basic introduction helps to adjust the knowledge of the reader accordingly. A practical treatment of the material enables a comprehensive understanding of how to handle specific problems, such as inhomogeneous illumination or occlusion. With this book, the reader should be able to develop object-oriented programs and show mathematical basic understanding. Such topics as image processing, navigation, camera types and camera calibration structure the described steps of developing further applications of vision-based robot projects.