GPU-ACCELERATED FORWARD-BACKWARD ALGORITHM WITH APPLICATION TO LATTICE-FREE MMI

被引：3
|
作者：
Ondel, Lucas [1 ]
Lam-Yee-Mui, Lea-Marie [1 ,3 ]
Kocour, Martin [2 ]
Corro, Caio Filippo [1 ]
Burget, Lukas [2 ]
机构：
[1] Univ Paris Saclay, LISN, CNRS, Orsay, France
[2] Brno Univ Technol, Fac Informat Technol, Brno, Czech Republic
[3] Vocapia Res, Orsay, France
来源：
2022 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP) | 2022年
关键词：
Lattice-Free MMI; end-to-end ASR; Julia language; forward-backward;
D O I：
10.1109/ICASSP43922.2022.9746824
中图分类号：
O42 [声学];
学科分类号：
070206 ; 082403 ;
摘要：
We propose to express the forward-backward algorithm in terms of operations between sparse matrices in a specific semiring. This new perspective naturally leads to a GPU-friendly algorithm which is easy to implement in Julia or any programming languages with native support of semiring algebra. We use this new implementation to train a TDNN with the LF-MMI objective function and we compare the training time of our system with PyChain-a recently introduced C++/CUDA implementation of the LF-MMI loss. Our implementation is about two times faster while not having to use any approximation such as the "leaky-HMM".
引用
收藏
页码：8417 / 8421
页数：5
相关论文
共 50 条
  • [31] An accelerated forward-backward algorithm with a new linesearch for convex minimization problems and its applications
    Hanjing, Adisak
    Jailoka, Pachara
    Suantai, Suthep
    AIMS MATHEMATICS, 2021, 6 (06): : 6180 - 6200
  • [32] An accelerated viscosity forward-backward splitting algorithm with the linesearch process for convex minimization problems
    Suthep Suantai
    Pachara Jailoka
    Adisak Hanjing
    Journal of Inequalities and Applications, 2021
  • [33] An accelerated viscosity forward-backward splitting algorithm with the linesearch process for convex minimization problems
    Suantai, Suthep
    Jailoka, Pachara
    Hanjing, Adisak
    JOURNAL OF INEQUALITIES AND APPLICATIONS, 2021, 2021 (01)
  • [34] CONVERGENCE RATE ANALYSIS OF ACCELERATED FORWARD-BACKWARD ALGORITHM WITH GENERALIZED NESTEROV MOMENTUM SCHEME
    Lin, Yizun
    Li, Si
    Zhang, Yunzhong
    INTERNATIONAL JOURNAL OF NUMERICAL ANALYSIS AND MODELING, 2023, 20 (04) : 518 - 537
  • [35] Accelerating Forward-Backward Sweep Power Flow Computation on the GPU
    Shah, Saumya
    Zarghami, Mahyar
    Muyan-Ozcelik, Pinar
    49TH INTERNATIONAL CONFERENCE ON PARALLEL PROCESSING WORKSHOP PROCEEDINGS, ICPP 2020, 2020,
  • [36] GPU-Accelerated Population Annealing Algorithm: Frustrated Ising Antiferromagnet on the Stacked Triangular Lattice
    Borovsky, Michal
    Weigel, Martin
    Barash, Lev Yu.
    Zukovic, Milan
    MATHEMATICAL MODELING AND COMPUTATIONAL PHYSICS (MMCP 2015), 2016, 108
  • [37] An Inertial Forward-Backward Algorithm for Monotone Inclusions
    Lorenz, Dirk A.
    Pock, Thomas
    JOURNAL OF MATHEMATICAL IMAGING AND VISION, 2015, 51 (02) : 311 - 325
  • [38] An Inertial Forward-Backward Algorithm for Monotone Inclusions
    Dirk A. Lorenz
    Thomas Pock
    Journal of Mathematical Imaging and Vision, 2015, 51 : 311 - 325
  • [39] GPU-accelerated phase extraction algorithm for interferograms: A real-time application
    Zhu, Xiaoqiang
    Wu, Yongqian
    Liu, Fengwei
    OPTICAL METROLOGY AND INSPECTION FOR INDUSTRIAL APPLICATIONS IV, 2016, 10023
  • [40] Bayesian Recurrent Units and the Forward-Backward Algorithm
    Bittar, Alexandre
    Garner, Philip N.
    INTERSPEECH 2022, 2022, : 4137 - 4141
12345