A benchmark of selected algorithmic differentiation tools on some problems in computer vision and machine learning

DOI10.1080/10556788.2018.1435651zbMath1453.65050arXiv1807.10129OpenAlexW2793476006MaRDI QIDQ4685586

Filip Srajer, Andrew W. Fitzgibbon, Zuzana Kukelova

Publication date: 9 October 2018

Published in: Optimization Methods and Software (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/1807.10129

zbMATH Keywords

computer visionautomatic differentiationmachine learningbenchmark

Mathematics Subject Classification ID

Learning and adaptive systems in artificial intelligence (68T05) Numerical differentiation (65D25) Computer aspects of numerical algorithms (65Y99) Computational issues in computer and robotic vision (65D19)

Related Items

Automatic adjoint differentiation for gradient descent and model calibration, Computation of higher order Lie derivatives on the infinity computer, Unnamed Item, A Simulink-based software solution using the infinity computer methodology for higher order differentiation, Unnamed Item, Julia language in machine learning: algorithms, applications, and open issues

Uses Software

• ADOL-C
• Matlab
• MuPAD
• ADIC
• Eigen
• Theano
• TAPENADE
• ColPack
• Symbolic Math Toolbox
• ADIC2
• MonoSLAM
• ADiGator
• Ceres Solver
• Autograd
• clad
• F#

Cites Work

• Comparison of automatic and symbolic differentiation in mathematical modeling and computer simulation of rigid body systems
• An efficient overloaded method for computing derivatives of mathematical functions in MATLAB
• The Tapenade automatic differentiation tool
• ColPack
• Evaluating Derivatives
• Multiple View Geometry in Computer Vision
• Fast Reverse-Mode Automatic Differentiation using Expression Templates in C++