Matthew Moore

Mountain View, California, United States Contact Info
8K followers 500+ connections

Join to view profile

About

My goal is to make a dent in the universe.

I'm a highly technical product leader &…

Articles by Matthew

See all articles

Activity

Join now to see all activity

Experience & Education

  • Graphlan

View Matthew’s full experience

See their title, tenure and more.

or

By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.

Publications

  • ClearGrasp: 3D Shape Estimation of Transparent Objects for Manipulation

    ArXiv

    Transparent objects are a common part of everyday life, yet they possess unique visual properties that make them incredibly difficult for standard 3D sensors to produce accurate depth estimates for. In many cases, they often appear as noisy or distorted approximations of the surfaces that lie behind them. To address these challenges, we present ClearGrasp -- a deep learning approach for estimating accurate 3D geometry of transparent objects from a single RGB-D image for robotic manipulation…

    Transparent objects are a common part of everyday life, yet they possess unique visual properties that make them incredibly difficult for standard 3D sensors to produce accurate depth estimates for. In many cases, they often appear as noisy or distorted approximations of the surfaces that lie behind them. To address these challenges, we present ClearGrasp -- a deep learning approach for estimating accurate 3D geometry of transparent objects from a single RGB-D image for robotic manipulation. Given a single RGB-D image of transparent objects, ClearGrasp uses deep convolutional networks to infer surface normals, masks of transparent surfaces, and occlusion boundaries. It then uses these outputs to refine the initial depth estimates for all transparent surfaces in the scene. To train and test ClearGrasp, we construct a large-scale synthetic dataset of over 50,000 RGB-D images, as well as a real-world test benchmark with 286 RGB-D images of transparent objects and their ground truth geometries. The experiments demonstrate that ClearGrasp is substantially better than monocular depth estimation baselines and is capable of generalizing to real-world images and novel objects. We also demonstrate that ClearGrasp can be applied out-of-the-box to improve grasping algorithms' performance on transparent objects. Code, data, and benchmarks will be released.

    Other authors
    See publication

Patents

  • System and method for representing user processes as software packages in a software package management system

    Issued US US8490082

    A system, method, and computer program product for representing at least one of a user process and a system process as a software package in a software package management system. The method includes creating at least one package comprising at least one of a user process and a system process. The at least one package comprises a structure consistent with a software package so that the at least one package is able to be managed by a software package management system. The at least one package is…

    A system, method, and computer program product for representing at least one of a user process and a system process as a software package in a software package management system. The method includes creating at least one package comprising at least one of a user process and a system process. The at least one package comprises a structure consistent with a software package so that the at least one package is able to be managed by a software package management system. The at least one package is stored on an information system via the software package management system.

    Other inventors
    See patent
  • Method and apparatus for scheduling jobs on a network

    Issued US US7831972

    An apparatus and method for scheduling a job process on at least one node in a server data processing network. The scheduling is based on a state of at least two nodes in the network as well as a cost metric of installing a required application on at least one of the nodes in the network of nodes. The apparatus and method gathers metrics indicating properties associated with at least two nodes on the network, the metrics including a presence of an application necessary for running the process…

    An apparatus and method for scheduling a job process on at least one node in a server data processing network. The scheduling is based on a state of at least two nodes in the network as well as a cost metric of installing a required application on at least one of the nodes in the network of nodes. The apparatus and method gathers metrics indicating properties associated with at least two nodes on the network, the metrics including a presence of an application necessary for running the process. A first cost factor for scheduling the process on a node in the network having the application necessary for running the process is compared with a second cost factor for scheduling the process on a node in the network that does not have the application necessary for running the process. Finally, a node in the network is selected for scheduling the process.

    Other inventors
    See patent
  • Method and Apparatus for In-Application Deals

    Filed US US20130006743 A1

    A computer-implemented method for in-application advertising display on a second computer using a first computer. A request is received on a first computer to display an advertisement on a second computer, the request from the second computer and including an identification of an application and a location information. The first computer selects an advertisement containing a deal comprised of a discount for prepurchase from a first company and a reward from a second company. The second company…

    A computer-implemented method for in-application advertising display on a second computer using a first computer. A request is received on a first computer to display an advertisement on a second computer, the request from the second computer and including an identification of an application and a location information. The first computer selects an advertisement containing a deal comprised of a discount for prepurchase from a first company and a reward from a second company. The second company is selected based on the identification of the application and the selection is based on both the identification of the application and the location information. The first computer transmits the advertisement to the second computer. The first computer receives a signal from the second computer indicating a purchase of the deal by a user of the second computer.

    See patent

Recommendations received

More activity by Matthew

View Matthew’s full profile

  • See who you know in common
  • Get introduced
  • Contact Matthew directly
Join to view full profile

People also viewed

Explore collaborative articles

We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.

Explore More

Others named Matthew Moore in United States

Add new skills with these courses