The Experts below are selected from a list of 10446 Experts worldwide ranked by ideXlab platform

Bjorn Schuller - One of the best experts on this subject based on the ideXlab platform.

  • introducing currennt the munich open source cuda recurrent neural network toolkit
    Journal of Machine Learning Research, 2015
    Co-Authors: Felix Weninger, Johannes Bergmann, Bjorn Schuller
    Abstract:

    In this article, we introduce CURRENNT, an open-source parallel implementation of deep recurrent neural networks (RNNs) supporting graphics processing units (GPUs) through NVIDIA's Computed Unified Device Architecture (CUDA). CURRENNT supports uni- and bidirectional RNNs with Long Short-Term Memory (LSTM) memory cells which overcome the vanishing gradient problem. To our knowledge, CURRENNT is the first publicly available parallel implementation of deep LSTM-RNNs. Benchmarks are given on a noisy speech recognition task from the 2013 2nd CHiME Speech Separation and Recognition Challenge, where LSTM-RNNs have been shown to deliver best performance. In the result, double digit speedups in bidirectional LSTM training are achieved with respect to a reference single-threaded CPU implementation. CURRENNT is available under the GNU general public license from http://sourceforge.net/p/currennt.

  • introducing currennt the munich open source cuda recurrent neural network toolkit
    Journal of Machine Learning Research, 2015
    Co-Authors: Felix Weninger, Johannes Bergmann, Bjorn Schuller
    Abstract:

    In this article, we introduce CURRENNT, an open-source parallel implementation of deep recurrent neural networks (RNNs) supporting graphics processing units (GPUs) through NVIDIA's Computed Unified Device Architecture (CUDA). CURRENNT supports uni- and bidirectional RNNs with Long Short-Term Memory (LSTM) memory cells which overcome the vanishing gradient problem. To our knowledge, CURRENNT is the first publicly available parallel implementation of deep LSTM-RNNs. Benchmarks are given on a noisy speech recognition task from the 2013 2nd CHiME Speech Separation and Recognition Challenge, where LSTM-RNNs have been shown to deliver best performance. In the result, double digit speedups in bidirectional LSTM training are achieved with respect to a reference single-threaded CPU implementation. CURRENNT is available under the GNU general public license from http://sourceforge.net/p/currennt.

Felix Weninger - One of the best experts on this subject based on the ideXlab platform.

  • introducing currennt the munich open source cuda recurrent neural network toolkit
    Journal of Machine Learning Research, 2015
    Co-Authors: Felix Weninger, Johannes Bergmann, Bjorn Schuller
    Abstract:

    In this article, we introduce CURRENNT, an open-source parallel implementation of deep recurrent neural networks (RNNs) supporting graphics processing units (GPUs) through NVIDIA's Computed Unified Device Architecture (CUDA). CURRENNT supports uni- and bidirectional RNNs with Long Short-Term Memory (LSTM) memory cells which overcome the vanishing gradient problem. To our knowledge, CURRENNT is the first publicly available parallel implementation of deep LSTM-RNNs. Benchmarks are given on a noisy speech recognition task from the 2013 2nd CHiME Speech Separation and Recognition Challenge, where LSTM-RNNs have been shown to deliver best performance. In the result, double digit speedups in bidirectional LSTM training are achieved with respect to a reference single-threaded CPU implementation. CURRENNT is available under the GNU general public license from http://sourceforge.net/p/currennt.

  • introducing currennt the munich open source cuda recurrent neural network toolkit
    Journal of Machine Learning Research, 2015
    Co-Authors: Felix Weninger, Johannes Bergmann, Bjorn Schuller
    Abstract:

    In this article, we introduce CURRENNT, an open-source parallel implementation of deep recurrent neural networks (RNNs) supporting graphics processing units (GPUs) through NVIDIA's Computed Unified Device Architecture (CUDA). CURRENNT supports uni- and bidirectional RNNs with Long Short-Term Memory (LSTM) memory cells which overcome the vanishing gradient problem. To our knowledge, CURRENNT is the first publicly available parallel implementation of deep LSTM-RNNs. Benchmarks are given on a noisy speech recognition task from the 2013 2nd CHiME Speech Separation and Recognition Challenge, where LSTM-RNNs have been shown to deliver best performance. In the result, double digit speedups in bidirectional LSTM training are achieved with respect to a reference single-threaded CPU implementation. CURRENNT is available under the GNU general public license from http://sourceforge.net/p/currennt.

Douglas A Hass - One of the best experts on this subject based on the ideXlab platform.

  • the gentlemen s agreement soldiers on linux under gnu general public license versions 2 and 3
    Social Science Research Network, 2009
    Co-Authors: Douglas A Hass
    Abstract:

    Over the past five years, the open source community has continued its robust debate on the intellectual property issues surrounding the GNU general public license (GPL) and its most successful project, the collaboratively developed Linux operating system. Commercial and non-commercial members of the Linux community have evolved Linux's open source software development model to accommodate realities of copyright law and the need to secure both significant commercial participation and widespread industry adoption. The resulting "Gentlemen's Agreement" is still fragile, especially with the 2007 release of a new version of the GPL. Legal practitioners and commentators have the opportunity to help the Linux community strengthen the gentlemen's agreement by explaining its utility in both legal and technical arenas.Arguments by strong supporters of the broadest interpretation of the GPL that overstate the effect of the license serve both to deny the existence of this Gentlemen's Agreement and risk undermining the commercial/open source balance that has served Linux well for more than a decade. This paper outlines the GPL's legal shortcomings and boundaries to help bolster the community's practical, and undervalued, gentlemen's agreement that enables commercial participation in open source projects.

  • a gentlemen s agreement assessing the gnu general public license and its adaptation to linux
    Social Science Research Network, 2006
    Co-Authors: Douglas A Hass
    Abstract:

    The open source community is conducting a robust debate on the intellectual property issues surrounding the GNU general public license (GPL) a popular modified public domain software license, and Linux, its most successful project to date. The Linux community has evolved its open source development model to accommodate realities of copyright law and the need to secure both significant commercial participation and widespread industry adoption. The legal issues underlying this transformation have not undergone the same robust analysis. This paper sheds light on those issues and tests some of their limits.The GPL fails to define fundamental terms adequately, including the inconsistent use of based on (derivative works), the lack of a choice of law provision, and the ambiguous treatment of patents. The GPL holds itself out as a viral license, purporting to foist itself on any software developer who has incorporated GPL code into a project. These and other factors combined with the Linux community's outdated views on copyright protection for kernel modules make it unlikely that a court could give full effect to the GPL or protect open source code from closed source intrusions. The GPL, however, does act as the most important beacon for Linux and the rest of the open source world. Its most significant contribution may differ greatly from the one envisioned by its creators: collaborative, decentralized development rather than free software. Contrary to some non-legal analyses, the gentlemen's agreement model employed by Linux to ensure that both closed source and open source software can coexist is a legally defensible, common sense adaptation of the GPL.

R Sander - One of the best experts on this subject based on the ideXlab platform.

  • the community atmospheric chemistry box model caaba mecca 4 0
    Geoscientific Model Development, 2019
    Co-Authors: R Sander, Patrick Jockel, A J G Baumgaertner, Sergey Gromov, H Harder, David Cabreraperez, Franziska Frank, J U Groos, Vincent Huijnen, Vlassis A Karydis
    Abstract:

    Abstract. We present version 4.0 of the atmospheric chemistry box model CAABA/MECCA that now includes a number of new features: (i) skeletal mechanism reduction, (ii) the Mainz Organic Mechanism (MOM) chemical mechanism for volatile organic compounds, (iii) an option to include reactions from the Master Chemical Mechanism (MCM) and other chemical mechanisms, (iv) updated isotope tagging, and (v) improved and new photolysis modules (JVAL, RADJIMT, DISSOC). Further, when MECCA is connected to a global model, the new feature of coexisting multiple chemistry mechanisms (PolyMECCA/CHEMGLUE) can be used. Additional changes have been implemented to make the code more user-friendly and to facilitate the analysis of the model results. Like earlier versions, CAABA/MECCA-4.0 is a community model published under the GNU general public license.

  • the atmospheric chemistry box model caaba mecca 3 0
    Geoscientific Model Development, 2011
    Co-Authors: R Sander, Astrid Kerkweg, Patrick Jockel, A J G Baumgaertner, Sergey Gromov, H Harder, Dagmar Kubistin, E Regelin, Hella Riede, Adrian Sandu
    Abstract:

    Abstract. We present version 3.0 of the atmospheric chemistry box model CAABA/MECCA. In addition to a complete update of the rate coefficients to the most recent recommendations, a number of new features have been added: chemistry in multiple aerosol size bins; automatic multiple simulations reaching steady-state conditions; Monte-Carlo simulations with randomly varied rate coefficients within their experimental uncertainties; calculations along Lagrangian trajectories; mercury chemistry; more detailed isoprene chemistry; tagging of isotopically labeled species. Further changes have been implemented to make the code more user-friendly and to facilitate the analysis of the model results. Like earlier versions, CAABA/MECCA-3.0 is a community model published under the GNU general public license.

Johannes Bergmann - One of the best experts on this subject based on the ideXlab platform.

  • introducing currennt the munich open source cuda recurrent neural network toolkit
    Journal of Machine Learning Research, 2015
    Co-Authors: Felix Weninger, Johannes Bergmann, Bjorn Schuller
    Abstract:

    In this article, we introduce CURRENNT, an open-source parallel implementation of deep recurrent neural networks (RNNs) supporting graphics processing units (GPUs) through NVIDIA's Computed Unified Device Architecture (CUDA). CURRENNT supports uni- and bidirectional RNNs with Long Short-Term Memory (LSTM) memory cells which overcome the vanishing gradient problem. To our knowledge, CURRENNT is the first publicly available parallel implementation of deep LSTM-RNNs. Benchmarks are given on a noisy speech recognition task from the 2013 2nd CHiME Speech Separation and Recognition Challenge, where LSTM-RNNs have been shown to deliver best performance. In the result, double digit speedups in bidirectional LSTM training are achieved with respect to a reference single-threaded CPU implementation. CURRENNT is available under the GNU general public license from http://sourceforge.net/p/currennt.

  • introducing currennt the munich open source cuda recurrent neural network toolkit
    Journal of Machine Learning Research, 2015
    Co-Authors: Felix Weninger, Johannes Bergmann, Bjorn Schuller
    Abstract:

    In this article, we introduce CURRENNT, an open-source parallel implementation of deep recurrent neural networks (RNNs) supporting graphics processing units (GPUs) through NVIDIA's Computed Unified Device Architecture (CUDA). CURRENNT supports uni- and bidirectional RNNs with Long Short-Term Memory (LSTM) memory cells which overcome the vanishing gradient problem. To our knowledge, CURRENNT is the first publicly available parallel implementation of deep LSTM-RNNs. Benchmarks are given on a noisy speech recognition task from the 2013 2nd CHiME Speech Separation and Recognition Challenge, where LSTM-RNNs have been shown to deliver best performance. In the result, double digit speedups in bidirectional LSTM training are achieved with respect to a reference single-threaded CPU implementation. CURRENNT is available under the GNU general public license from http://sourceforge.net/p/currennt.