Hybrid Bayesian Eigenobjects: Combining Linear Subspace and Deep Network Methods for 3D Robot Vision

Burchfiel, Benjamin; Konidaris, George

doi:10.1109/iros.2018.8593795

Cited by 6 publications

(20 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Such a hybrid model where the deep learning is assisted by additional sensor sources like synthetic aperture radar (SAR) imagery and elevation like synthetic aperture radar (SAR) imagery and elevation is presented by [40]. In the context of 3D robot vision, [42] have shown that combining both linear subspace methods and deep convolutional prediction achieves improved performance along with several orders of magnitude faster runtime performance compared to the state of the art.…”

Section: Making Best Use Of Edge Computingmentioning

confidence: 99%

Deep Learning vs. Traditional Computer Vision

Mahony

Campbell

Carvalho

et al. 2019

Advances in Intelligent Systems and Computing

706

383

View full text Add to dashboard Cite

Deep Learning has pushed the limits of what was possible in the domain of Digital Image Processing. However, that is not to say that the traditional computer vision techniques which had been undergoing progressive development in years prior to the rise of DL have become obsolete. This paper will analyse the benefits and drawbacks of each approach. The aim of this paper is to promote a discussion on whether knowledge of classical computer vision techniques should be maintained. The paper will also explore how the two sides of computer vision can be combined. Several recent hybrid methodologies are reviewed which have demonstrated the ability to improve computer vision performance and to tackle problems not suited to Deep Learning. For example, combining traditional computer vision techniques with Deep Learning has been popular in emerging domains such as Panoramic Vision and 3D vision for which Deep Learning models have not yet been fully optimised.

show abstract

Section: Making Best Use Of Edge Computingmentioning

confidence: 99%

Deep Learning vs. Traditional Computer Vision

Mahony

Campbell

Carvalho

et al. 2019

Advances in Intelligent Systems and Computing

706

383

View full text Add to dashboard Cite

show abstract

“…Bayesian Eigenobjects (BEOs) [3], [17] offer compact representations of objects. Using 3D object models for training, BEOs generate a low-dimensional subspace serving as a basis which well captures the object classes used to train it.…”

Section: Background and Related Workmentioning

confidence: 99%

“…This makes the BEO subspace a compelling target for language mapping in a robotics domain. Critically, the hybrid variant of BEOs (HBEOs) [17] learn an explicit subspace like BEOs, but use a deep convolutional network to learn an embedding directly from a depth image into the object-subspace, allowing for high performance, fast runtime, and the ability to complete objects without requiring voxelization of observed objects.…”

Section: Background and Related Workmentioning

confidence: 99%

“…The low-dimensional BEO embedding is learned from a set of 3D meshes from a single class. These meshes are aligned and voxelized as in Burchfiel and Konidaris [17]. Each voxelized object is then stacked into a vectorrepresentation; an object is now a point in high-dimensional voxel space.…”

Section: A Obtaining Low-dimensional Beo Embeddingsmentioning

confidence: 99%

“…More recently, work has progressed on learned models that predict 3D shape from partially-observed input [14], [15], [16], [3]. While most of these methods require known voxelization of the partially observed query object, recent work has allowed for prediction directly from a depth image, an important requirement for most robotic applications [17].…”

Section: Background and Related Workmentioning

confidence: 99%

See 2 more Smart Citations

Grounding Language Attributes to Objects using Bayesian Eigenobjects

Cohen

Burchfiel

Nguyen

et al. 2019

2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

Self Cite

View full text Add to dashboard Cite

We develop a system to disambiguate object instances within the same class based on simple physical descriptions. The system takes as input a natural language phrase and a depth image containing a segmented object and predicts how similar the observed object is to the object described by the phrase. Our system is designed to learn from only a small amount of human-labeled language data and generalize to viewpoints not represented in the language-annotated depth image training set. By decoupling 3D shape representation from language representation, this method is able to ground language to novel objects using a small amount of languageannotated depth-data and a larger corpus of unlabeled 3D object meshes, even when these objects are partially observed from unusual viewpoints. Our system is able to disambiguate between novel objects, observed via depth images, based on natural language descriptions. Our method also enables viewpoint transfer; trained on human-annotated data on a small set of depth images captured from frontal viewpoints, our system successfully predicted object attributes from rear views despite having no such depth images in its training set. Finally, we demonstrate our approach on a Baxter robot, enabling it to pick specific objects based on human-provided natural language descriptions.

show abstract

Advances in Computer Vision

Mahony¹,

Campbell²,

Carvalho³

et al. 2020

Advances in Intelligent Systems and Computing

View full text Add to dashboard Cite

The series "Advances in Intelligent Systems and Computing" contains publications on theory, applications, and design methods of Intelligent Systems and Intelligent Computing. Virtually all disciplines such as engineering, natural sciences, computer and information science, ICT, economics, business, e-commerce, environment, healthcare, life science are covered. The list of topics spans all the areas of modern intelligent systems and computing such as: computational intelligence, soft computing including neural networks, fuzzy systems, evolutionary computing and the fusion of these paradigms, social intelligence, ambient intelligence, computational neuroscience, artificial life, virtual worlds and society, cognitive science and systems, Perception and Vision, DNA and immune based systems, self-organizing and adaptive systems, e-Learning and teaching, human-centered and human-centric computing, recommender systems, intelligent control, robotics and mechatronics including human-machine teaming, knowledge-based paradigms, learning paradigms, machine ethics, intelligent data analysis, knowledge management, intelligent agents, intelligent decision making and support, intelligent network security, trust management, interactive entertainment, Web intelligence and multimedia.The publications within "Advances in Intelligent Systems and Computing" are primarily proceedings of important conferences, symposia and congresses. They cover significant recent developments in the field, both of a foundational and applicable character. An important characteristic feature of the series is the short publication time and world-wide distribution. This permits a rapid and broad dissemination of research results.

show abstract

Hybrid Bayesian Eigenobjects: Combining Linear Subspace and Deep Network Methods for 3D Robot Vision

Cited by 6 publications

References 30 publications

Deep Learning vs. Traditional Computer Vision

Deep Learning vs. Traditional Computer Vision

Grounding Language Attributes to Objects using Bayesian Eigenobjects

Advances in Computer Vision

Contact Info

Product

Resources

About