DOCSLIB.ORG

Appendices at Indices I and J Is Also Necessary When There Are Differ- Ent Numbers of Points Per Minibatch but Batched Together Using Zero Padding

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Appendices at Indices I and J Is Also Necessary When There Are Differ- Ent Numbers of Points Per Minibatch but Batched Together Using Zero Padding Generalizing Convolutional Neural Networks for Equivariance to Lie Groups on Arbitrary Continuous Data Appendices at indices i and j is also necessary when there are differ- ent numbers of points per minibatch but batched together using zero padding. The generalized PointConv trick can A. Derivations and Additional Methodology thus be applied in batch mode when there may be varied number of points per example and varied number of points A.1. Generalized PointConv Trick per neighborhood. The matrix notation becomes very cumbersome for manipu- lating these higher order n-dimensional arrays, so we will A.2. Abelian G and Coordinate Transforms instead use index notation with Latin indices i; j; k index- For Abelian groups that cover in a single orbit, the ing points, Greek indices α; β; γ indexing feature channels, computation is very similar to ordinaryX Euclidean convo- and c indexing the coordinate dimensions of which there lution. Defining ai = log(ui), bj = log(vj), and using are d = 3 for PointConv and d = dim(G) + 2 dim(Q) bj ai ai bj 1 the fact that e− e = e − means that log(v− ui) = for LieConv.3 As the objects are not geometric tensors but j (log exp)(a b ). Defining f~ = f exp, h~ = h exp; simply n-dimensional arrays, we will make no distinction i j we get◦ − ◦ ◦ between upper and lower indices. After expanding into in- dices, it should be assumed that all values are scalars, and 1 h~(a ) = (k~ proj)(a b )f~(b ); (15) that any free indices can range over all of the values. i n θ ◦ i − j j j nbhd(i) 2 X Let kα,β be the output of the MLP k which takes ac ij θ ij where proj = log exp projects to the image of the loga- as input and acts independently over the locations i; jf. Forg rithm map. Apart from◦ a projection and a change to logarith- PointConv, the input ac = xc xc and for LieConv the ij i j mic coordinates, this is equivalent to Euclidean convolution c 1 − c input aij = Concat([log(vj− ui); qi; qj]) . in a vector space with dimensionality of the group. When We wish to compute the group is Abelian and is a homogeneous space, then the dimension of the groupX is the dimension of the input. In α α,β β hi = kij fj : (12) these cases we have a trivial stabilizer group H and single origin o, so we can view f and h as acting on the input Xj,β xi = uio. In Wu et al.(2019), it was observed that since kα,β is the ij This directly generalizes some of the existing coordinate α,β α,β γ output of an MLP, kij = γ Wγ si;j for some final transform methods for achieving equivariance from the liter- weight matrix W and penultimate activations sγ (sγ is P i;j i;j ature such as log polar coordinates for rotation and scaling simply the result of the MLP after the last nonlinearity). equivariance (Esteves et al., 2017), and using hyperbolic With this in mind, we can rewrite (12) coordinates for squeeze and scaling equivariance. α α,β γ β Log Polar Coordinates: Consider the Abelian Lie group hi = Wγ si;j fj (13) j,β γ of positive scalings and rotations: G = R∗ SO(2) acting X X 2 × on R . Elements of the group M G can be expressed as α,β γ β (14) 2 = Wγ si;jfj a 2 2 matrix β,γ j × X X r cos(θ) r sin(θ) M(r; θ) = r sin(θ) −r cos(θ) In practice, the intermediate number of channels is much less than the product of c and c : γ < α β and + 4 in out for r R and θ R. The matrix logarithm is so this reordering of the computation leadsj j toj ajj massivej 2 2 reduction in both memory and compute. Furthermore, r cos(θ) r sin(θ) log(r) θ mod 2π log − = − ; bγ,β = sγ f β can be implemented with regular ma- r sin(θ) r cos(θ) θ mod 2π log(r) i j i;j j α α,β γ,β trix multiplication and hi = β,γ Wγ bi can be also or more compactly log(M(r; θ)) = log(r)I+(θ mod 2π)J, P α α,ε " by flattening (β; γ) into a single axis ": hi = " W bi . which is mod in the basis for the Lie algebra P [log(r); θ 2π] [I;J]. It is clear that proj = log exp is simply mod 2π The sum over index j can be restricted to a subsetPj(i) (such ◦ β on the J component. as a chosen neighborhood) by computing f( ) at each of the · required indices and padding to the size of the maximum As R2 is a homogeneous space of G, one can choose the γ,β γ β 2 subset with zeros, and computing bi = j si;j(i)fj(i) us- global origin o = [1; 0] R . A little algebra shows that ing dense matrix multiplication. Masking out of the values 2 P 4Here θ mod 2π is defined to mean θ + 2πn for the integer 3dim(Q) is the dimension of the space into which Q, the orbit n such that the value is in (−π; π), consistent with the principal identifiers, are embedded. matrix logarithm. (θ + π)%2π − π in programming notation. Generalizing Convolutional Neural Networks for Equivariance to Lie Groups on Arbitrary Continuous Data lifting to the group yields the transformation ui = M(ri; θi) A.3. Sufficient Conditions for Geodesic Distance 2 2 for each point pi = uio, where r = x + y , and 1 In general, the function d(u; v) = log(v− u) F , defined θ = atan2(y; x) are the polar coordinates of the point pi. k k 1 p on the domain of GL(d) covered by the exponential map, Observe that the logarithm of v− u has a simple expression j i satisfies the first three conditions of a distance metric but highlighting the fact that it is invariant to scale and rotational not the triangle inequality, making it a semi-metric: transformations of the elements, 1 1 log(v− u ) = log(M(r ; θ ) M(r ; θ )) 1. d(u; v) 0 j i j j − i i ≥ 1 = log(ri=rj)I + (θi θj mod 2π)J: 2. d(u; v) = 0 log(u− v) = 0 u = v − , , 1 1 3. d(u; v) = log(v− u) = log(u− v) = d(v; u). k k k − k Now writing out our Monte Carlo estimation of the integral: However for certain subgroups of GL(d) with additional 1 structure, the triangle inequality holds and the function is h(p ) = k~ (log(r =r ); θ θ mod 2π)f(p ); i n θ i j i − j j the distance along geodesics connecting group elements u j X and v according to the metric tensor which is a discretization of the log polar convolution from T T 1 A; B := Tr(A u− u− B); (16) Esteves et al.(2017). This can be trivially extended to h iu encompass cylindrical coordinates with the group T (1) where T denotes inverse and transpose. × − R∗ SO(2). × Specifically, if the subgroup G is in the image of the exp : Hyperbolic coordinates: For another nontrivial example, g G map and each infinitesmal generator commutes with ! T consider the group of scalings and squeezes G = R∗ SQ its transpose: [A; A ] = 0 for A g, then d(u; v) = 2 × 1 8 2 acting on the positive orthant = (x; y) R : x > log(v− u) F is the geodesic distance between u; v. 0; y > 0 . Elements of the groupX canf be expressed2 as the k k Geodesic Equation: product ofg a squeeze mapping and a scaling Geodesics of (16) satisfying γ_ γ_ = 0 can equivalently be derived by minimizing ther energy s 0 r 0 rs 0 functional M(r; s) = = 0 1=s 0 r 0 r=s 1 T T 1 E[γ] = γ;_ γ_ γ dt = Tr(_γ γ− γ− γ_ )dt for any r; s +. As the group is abelian, the logarithm h i R Zγ Z0 splits nicely2 in terms of the two generators I and A: using the calculus of variations. Minimizing curves γ(t), rs 0 1 0 1 0 connecting elements u and v in G (γ(0) = v; γ(1) = u) log = (log r) + (log s) : 0 r=s 0 1 0 1 satisfy − 1 Again is a homogeneous space of G, and we choose a T T 1 single originX o = [1; 1]. With a little algebra, it is clear that 0 = δE = δ Tr(_γ γ− γ− γ_ )dt Z0 M(ri; si)o = pi where r = pxy and s = x=y are the Noting that δ(γ 1) = γ 1δγγ 1 and the linearity of the hyperbolic coordinates of pi. − − − p trace, − Expressed in the basis = [I;A] for the Lie algebra above, B 1 we see that T T 1 T T 1 1 2 Tr(_γ γ− γ− δγ_ ) Tr(_γ γ− γ− δγγ− γ_ )dt = 0: 1 0 − log(vj− ui) = log(ri=rj)I + log(si=sj)A Z yielding the expression for convolution Using the cyclic property of the trace and integrating by parts, we have that 1 h(pi) = k~θ(log(ri=rj); log(si=sj))f(pj); 1 n d T T 1 1 T T 1 j 2 Tr (_γ γ− γ− )+γ− γ_ γ_ γ− γ− δγ dt = 0; X − dt Z0 which is equivariant to squeezes and scalings.
Load more

Recommended publications
  • Entropy Generation in Gaussian Quantum Transformations: Applying the Replica Method to Continuous-Variable Quantum Information Theory
    www.nature.com/npjqi All rights reserved 2056-6387/15 ARTICLE OPEN Entropy generation in Gaussian quantum transformations: applying the replica method to continuous-variable quantum information theory Christos N Gagatsos1, Alexandros I Karanikas2, Georgios Kordas2 and Nicolas J Cerf1 In spite of their simple description in terms of rotations or symplectic transformations in phase space, quadratic Hamiltonians such as those modelling the most common Gaussian operations on bosonic modes remain poorly understood in terms of entropy production. For instance, determining the quantum entropy generated by a Bogoliubov transformation is notably a hard problem, with generally no known analytical solution, while it is vital to the characterisation of quantum communication via bosonic channels. Here we overcome this difficulty by adapting the replica method, a tool borrowed from statistical physics and quantum field theory. We exhibit a first application of this method to continuous-variable quantum information theory, where it enables accessing entropies in an optical parametric amplifier. As an illustration, we determine the entropy generated by amplifying a binary superposition of the vacuum and a Fock state, which yields a surprisingly simple, yet unknown analytical expression. npj Quantum Information (2015) 2, 15008; doi:10.1038/npjqi.2015.8; published online 16 February 2016 INTRODUCTION Gaussian states (e.g., the vacuum state, resulting after amplifica- Gaussian transformations are ubiquitous in quantum physics, tion in a thermal state of well-known
    [Show full text]
  • TEC Forecasting Based on Manifold Trajectories
    Article TEC forecasting based on manifold trajectories Enrique Monte Moreno 1,* ID , Alberto García Rigo 2,3 ID , Manuel Hernández-Pajares 2,3 ID and Heng Yang 1 ID 1 Department of Signal Theory and Communications, TALP research center, Technical University of Catalonia, Barcelona, Spain 2 UPC-IonSAT, Technical University of Catalonia, Barcelona, Spain 3 IEEC-CTE-CRAE, Institut d’Estudis Espacials de Catalunya, Barcelona, Spain * Correspondence: [email protected]; Tel.: +34-934016435 Academic Editor: name Version June 8, 2018 submitted to Remote Sens. 1 Abstract: In this paper, we present a method for forecasting the ionospheric Total Electron Content 2 (TEC) distribution from the International GNSS Service’s Global Ionospheric Maps. The forecasting 3 system gives an estimation of the value of the TEC distribution based on linear combination of 4 previous TEC maps (i.e. a set of 2D arrays indexed by time), and the computation of a tangent 5 subspace in a manifold associated to each map. The use of the tangent space to each map is justified 6 because it allows to model the possible distortions from one observation to the next as a trajectory on 7 the tangent manifold of the map. The coefficients of the linear combination of the last observations 8 along with the tangent space are estimated at each time stamp in order to minimize the mean square 9 forecasting error with a regularization term. The the estimation is made at each time stamp to adapt 10 the forecast to short-term variations in solar activity.. 11 Keywords: Total Electron Content; Ionosphere; Forecasting; Tangent Distance; GNSS 12 1.
    [Show full text]
  • MATRICES Part 2 3. Linear Equations
    MATRICES part 2 (modified content from Wikipedia articles on matrices http://en.wikipedia.org/wiki/Matrix_(mathematics)) 3. Linear equations Matrices can be used to compactly write and work with systems of linear equations. For example, if A is an m-by-n matrix, x designates a column vector (i.e., n×1-matrix) of n variables x1, x2, ..., xn, and b is an m×1-column vector, then the matrix equation Ax = b is equivalent to the system of linear equations a1,1x1 + a1,2x2 + ... + a1,nxn = b1 ... am,1x1 + am,2x2 + ... + am,nxn = bm . For example, 3x1 + 2x2 – x3 = 1 2x1 – 2x2 + 4x3 = – 2 – x1 + 1/2x2 – x3 = 0 is a system of three equations in the three variables x1, x2, and x3. This can be written in the matrix form Ax = b where 3 2 1 1 1 A = 2 2 4 , x = 2 , b = 2 1 1/2 −1 푥3 0 � − � �푥 � �− � A solution to− a linear system− is an assignment푥 of numbers to the variables such that all the equations are simultaneously satisfied. A solution to the system above is given by x1 = 1 x2 = – 2 x3 = – 2 since it makes all three equations valid. A linear system may behave in any one of three possible ways: 1. The system has infinitely many solutions. 2. The system has a single unique solution. 3. The system has no solution. 3.1. Solving linear equations There are several algorithms for solving a system of linear equations. Elimination of variables The simplest method for solving a system of linear equations is to repeatedly eliminate variables.
    [Show full text]
  • TEC Forecasting Based on Manifold Trajectories
    remote sensing Article TEC Forecasting Based on Manifold Trajectories Enrique Monte Moreno 1,* ID , Alberto García Rigo 2,3 ID , Manuel Hernández-Pajares 2,3 ID and Heng Yang 1 ID 1 TALP Research Center, Department of Signal Theory and Communications, Polytechnical University of Catalonia, 08034 Barcelona, Spain; [email protected] 2 UPC-IonSAT, Polytechnical University of Catalonia, 08034 Barcelona, Spain; [email protected] (A.G.R.); [email protected] (M.H.-P.) 3 IEEC-CTE-CRAE, Institut d’Estudis Espacials de Catalunya, 08034 Barcelona, Spain * Correspondence: [email protected]; Tel.: +34-934016435 Received: 18 May 2018; Accepted: 13 June 2018; Published: 21 June 2018 Abstract: In this paper, we present a method for forecasting the ionospheric Total Electron Content (TEC) distribution from the International GNSS Service’s Global Ionospheric Maps. The forecasting system gives an estimation of the value of the TEC distribution based on linear combination of previous TEC maps (i.e., a set of 2D arrays indexed by time), and the computation of a tangent subspace in a manifold associated to each map. The use of the tangent space to each map is justified because it allows modeling the possible distortions from one observation to the next as a trajectory on the tangent manifold of the map. The coefficients of the linear combination of the last observations along with the tangent space are estimated at each time stamp to minimize the mean square forecasting error with a regularization term. The estimation is made at each time stamp to adapt the forecast to short-term variations in solar activity.
    [Show full text]
  • Geometric Analysis: Partial Differential Equations and Surfaces
    570 Geometric Analysis: Partial Differential Equations and Surfaces UIMP-RSME Lluis Santaló Summer School 2010: Geometric Analysis June 28–July 2, 2010 University of Granada, Spain Joaquín Pérez José A. Gálvez Editors American Mathematical Society Real Sociedad Matemática Española American Mathematical Society Geometric Analysis: Partial Differential Equations and Surfaces UIMP-RSME Lluis Santaló Summer School 2010: Geometric Analysis June 28–July 2, 2010 University of Granada, Spain Joaquín Pérez José A. Gálvez Editors 570 Geometric Analysis: Partial Differential Equations and Surfaces UIMP-RSME Lluis Santaló Summer School 2010: Geometric Analysis June 28–July 2, 2010 University of Granada, Spain Joaquín Pérez José A. Gálvez Editors American Mathematical Society Real Sociedad Matemática Española American Mathematical Society Providence, Rhode Island EDITORIAL COMMITTEE Dennis DeTurck, Managing Editor George Andrews Abel Klein Martin J. Strauss Editorial Committee of the Real Sociedad Matem´atica Espa˜nola Pedro J. Pa´ul, Director Luis Al´ıas Emilio Carrizosa Bernardo Cascales Javier Duoandikoetxea Alberto Elduque Rosa Maria Mir´o Pablo Pedregal Juan Soler 2010 Mathematics Subject Classification. Primary 53A10; Secondary 35B33, 35B40, 35J20, 35J25, 35J60, 35J96, 49Q05, 53C42, 53C45. Library of Congress Cataloging-in-Publication Data UIMP-RSME Santal´o Summer School (2010 : University of Granada) Geometric analysis : partial differential equations and surfaces : UIMP-RSME Santal´o Summer School geometric analysis, June 28–July 2, 2010, University of Granada, Granada, Spain / Joaqu´ın P´erez, Jos´eA.G´alvez, editors p. cm. — (Contemporary Mathematics ; v. 570) Includes bibliographical references. ISBN 978-0-8218-4992-7 (alk. paper) 1. Minimal surfaces–Congresses. 2. Geometry, Differential–Congresses. 3. Differential equations, Partial–Asymptotic theory–Congresses.
    [Show full text]
  • Generalized Multiple Correlation Coefficient As a Similarity
    Generalized Multiple Correlation Coefficient as a Similarity Measurement between Trajectories Julen Urain1 Jan Peters1;2 Abstract— Similarity distance measure between two trajecto- ries is an essential tool to understand patterns in motion, for example, in Human-Robot Interaction or Imitation Learning. The problem has been faced in many fields, from Signal Pro- cessing, Probabilistic Theory field, Topology field or Statistics field. Anyway, up to now, none of the trajectory similarity measurements metrics are invariant to all possible linear trans- formation of the trajectories (rotation, scaling, reflection, shear mapping or squeeze mapping). Also not all of them are robust in front of noisy signals or fast enough for real-time trajectory classification. To overcome this limitation this paper proposes a similarity distance metric that will remain invariant in front of any possible linear transformation. Based on Pearson’s Correlation Coefficient and the Coefficient of Determination, our similarity metric, the Generalized Multiple Correlation Fig. 1: GMCC applied for clustering taught trajectories in Coefficient (GMCC) is presented like the natural extension of Imitation Learning the Multiple Correlation Coefficient. The motivation of this paper is two-fold: First, to introduce a new correlation metric that presents the best properties to compute similarities between trajectories invariant to linear transformations and compare human-robot table tennis match. In [11], HMM were used it with some state of the art similarity distances. Second, to to recognize human gestures. The recognition was invariant present a natural way of integrating the similarity metric in an to the starting position of the gestures. The sequences of Imitation Learning scenario for clustering robot trajectories.
    [Show full text]
  • Group Quantization of Quadratic Hamiltonians in Finance
    Date: February 2 2021 Version 1.0 Group Quantization of Quadratic Hamiltonians in Finance Santiago Garc´ıa1 Abstract The Group Quantization formalism is a scheme for constructing a functional space that is an irreducible infinite dimensional representation of the Lie algebra belonging to a dynamical symmetry group. We apply this formalism to the construction of functional space and operators for quadratic potentials- gaussian pricing kernels in finance. We describe the Black-Scholes theory, the Ho-Lee interest rate model and the Euclidean repulsive and attractive oscillators. The symmetry group used in this work has the structure of a principal bundle with base (dynamical) group a + semi-direct extension of the Heisenberg-Weyl group by SL(2;R), and structure group (fiber) R . + By using a R central extension, we obtain the appropriate commutator between the momentum and coordinate operators [pˆ;xˆ] = 1 from the beginning, rather than the quantum-mechanical [pˆ;xˆ] = −i}. The integral transforma- tion between momentum and coordinate representations is the bilateral Laplace transform, an integral transform associated to the symmetry group. Keywords Black-Scholes – Lie Groups – Quadratic Hamiltonians – Central Extensions – Oscillator – Linear Canonical Transformations – Mellin Transform – Laplace Transform [email protected] Contents Introduction 1 Galilean Transformations as Numeraire Change3 The Group Quantization formalism ([1], [2], [3], [6]) is a 2 The Group Quantization Formalism3 scheme for constructing a functional space that is an irre- ducible infinite dimensional representation of the Lie algebra 3 The WSp(2;R) Group6 belonging to a dynamical symmetry group. 4 Black-Scholes9 This formalism utilizes Cartan geometries in a framework 5 Linear Potential.
    [Show full text]
  • {PDF} Matrices and Linear Transformations Ebook Free Download
    MATRICES AND LINEAR TRANSFORMATIONS PDF, EPUB, EBOOK Charles G. Cullen | 336 pages | 07 Jan 1991 | Dover Publications Inc. | 9780486663289 | English | New York, United States Matrices and Linear Transformations PDF Book Parallel projections are also linear transformations and can be represented simply by a matrix. Further information: Orthogonal projection. Preimage and kernel example Opens a modal. Linear transformation examples. Matrix vector products as linear transformations Opens a modal. Rotation in R3 around the x-axis Opens a modal. So what is a1 plus b? That's my first condition for this to be a linear transformation. Let me define my transformation. This means that applying the transformation T to a vector is the same as multiplying by this matrix. Distributive property of matrix products Opens a modal. Sorry, what is vector a plus vector b? The distinction between active and passive transformations is important. Name optional. Now, we just showed you that if I take the transformations separately of each of the vectors and then add them up, I get the exact same thing as if I took the vectors and added them up first and then took the transformation. Example of finding matrix inverse Opens a modal. We can view it as factoring out the c. We already had linear combinations so we might as well have a linear transformation. Part of my definition I'm going to tell you, it maps from r2 to r2. In some practical applications, inversion can be computed using general inversion algorithms or by performing inverse operations that have obvious geometric interpretation, like rotating in opposite direction and then composing them in reverse order.
    [Show full text]
  • Affine Shape Descriptor Algorithm for Image Matching and Object Recognition in a Digital Image
    Revista Iberoamericana de las Ciencias Computacionales e Informática ISSN: 2007-9915 Affine Shape Descriptor Algorithm for Image Matching and Object Recognition in a Digital Image Algoritmo para clasificación de formas invariante a transformaciones afines para objetos rígidos en una imagen digital Algoritmo de classificação para as formas invariantes afins para objetos rígidos em uma imagem digital DOI: http://dx.doi.org/10.23913/reci.v6i11.57 Jacob Morales G. Centro Universitario de Ciencias Exactas e Ingenierías, Universidad de Guadalajara, México [email protected] Nancy Arana D. Centro Universitario de Ciencias Exactas e Ingenierías, Universidad de Guadalajara, México [email protected] Alberto A. Gallegos Centro Universitario de Ciencias Exactas e Ingenierías, Universidad de Guadalajara, México [email protected] Resumen El uso de la forma como criterio discriminante entre clases de objetos extraídos de una imagen digital, es uno de los roles más importantes en la visión por computadora. El uso de la forma ha sido estudiado extensivamente en décadas recientes, porque el borde de los objetos guarda suficiente información para su correcta clasificación, adicionalmente, la cantidad de memoria utilizada para guardar un borde, es mucho menor en comparación a la cantidad de información contenida en la región que delimita. En este artículo, se propone un nuevo descriptor de forma. Se demuestra que el descriptor es invariante a la perspectiva, es estable en la presencia de ruido, y puede diferenciar entre diferentes clases de objetos. Un análisis comparativo es incluido para mostrar el rendimiento de nuestra propuesta con respecto a los algoritmos del estado del arte. Vol. 6, Núm. 11 Enero - Junio 2017 RECI Revista Iberoamericana de las Ciencias Computacionales e Informática ISSN: 2007-9915 Palabras Clave: Reconocimiento de Patrones, Visión por Computadora, Descriptor de Forma, Transformación por Afinidad.
    [Show full text]
  • 3D Oriented Projective Geometry Through Versors of R3,3
    Adv. Appl. Clifford Algebras c The Author(s) This article is published with open access at Springerlink.com 2015 Advances in DOI 10.1007/s00006-015-0625-y Applied Clifford Algebras 3D Oriented Projective Geometry Through Versors of R3,3 Leo Dorst* Abstract. It is possible to set up a correspondence between 3D space and R3,3, interpretable as the space of oriented lines (and screws), such that special projective collineations of the 3D space become represented as rotors in the geometric algebra of R3,3. We show explicitly how various primitive projective transformations (translations, rotations, scalings, perspectivities, Lorentz transformations) are represented, in geomet- rically meaningful parameterizations of the rotors by their bivectors. Odd versors of this representation represent projective correlations, so (oriented) reflections can only be represented in a non-versor manner. Specifically, we show how a new and useful ‘oriented reflection’ can be defined directly on lines. We compare the resulting framework to the unoriented R3,3 approach of Klawitter (Adv Appl Clifford Algebra, 24:713–736, 2014), and the R4,4 rotor-based approach by Goldman et al. (Adv Appl Clifford Algebra, 25(1):113–149, 2015) in terms of expres- siveness and efficiency. Mathematics Subject Classification. 15A33, 51M35. Keywords. Projective geometry, Oriented projective geometry, Geometric algebra, Homogeneous coordinates, Pl¨ucker coordinates, Oriented lines, Projective collineation, Versor, Rotor, Bivector generator, Oriented reflection. 1. Introduction 1.1. Towards a Geometric Algebra of Projective Geometry Projective transformations in 3D and 2D are extensively used in computer vision and computer graphics. In those fields, it is common practice to con- struct a desired projective transformation from certain geometrically mean- ingful primitive operations (rotations at the origin, translations, scalings, skews, etc.), or to analyze a given transformation in terms of such primi- tive operations.
    [Show full text]
  • MATRICES Part 2 3. Linear Equations
    MATRICES part 2 (modified content from Wikipedia articles on matrices http://en.wikipedia.org/wiki/Matrix_(mathematics)) 3. Linear equations Matrices can be used to compactly write and work with systems of linear equations. For example, if A is an m-by-n matrix, x designates a column vector (i.e., n×1-matrix) of n variables x1, x2, ..., xn, and b is an m×1-column vector, then the matrix equation Ax = b is equivalent to the system of linear equations a1,1x1 + a1,2x2 + ... + a1,nxn = b1 ... am,1x1 + am,2x2 + ... + am,nxn = bm . For example, 3x1 + 2x2 – x3 = 1 2x1 – 2x2 + 4x3 = – 2 – x1 + 1/2x2 – x3 = 0 is a system of three equations in the three variables x1, x2, and x3. This can be written in the matrix form Ax = b where 3 2 1 1 1 A = 2 2 4 , x = 2 , b = 2 1 1/2 −1 푥3 0 � − � �푥 � �− � A solution to− a linear system− is an assignment푥 of numbers to the variables such that all the equations are simultaneously satisfied. A solution to the system above is given by x1 = 1 x2 = – 2 x3 = – 2 since it makes all three equations valid. A linear system may behave in any one of three possible ways: 1. The system has infinitely many solutions. 2. The system has a single unique solution. 3. The system has no solution. 3.1. Solving linear equations There are several algorithms for solving a system of linear equations. Elimination of variables The simplest method for solving a system of linear equations is to repeatedly eliminate variables.
    [Show full text]
  • Mathematical Tables and Algorithms
    Appendix A Mathematical Tables and Algorithms This appendix provides lists with some definitions, integrals, trigonometric relations and the like. It is not exhaustive, presenting only that information used throughout the book and a few more just for complementation purposes. The reader can find extensive tables available in the literature. We recommend the works by Abramowitz and Stegun [1], Bronshtein, Semendyayev, Musiol and Muehlig [2], Gradshteyn and Ryzhik [3], Polyanin and Manzhirov [4] and Zwillinger [7]. We also recommend the book by Poularikas [5], though its major application is in the signal processing area. Nevertheless, it complements the tables of the other recommended books. In terms of online resources, the Digital Library of Mathematical Functions (DLMF) is intended to be a reference data for special functions and their applications [6]. It is also intended to be an update of [1]. A.1 Trigonometric Relations ∞ x3 x5 x7 (−1)n x2n+1 sin x = x − + − +···= (A.1) 3! 5! 7! (2n + 1)! n=0 ∞ x2 x4 x6 (−1)n x2n cos x = 1 − + − +···= (A.2) 2! 4! 6! (2n)! n=0 1 sin2 θ = (1 − cos 2θ)(A.3) 2 1 cos2 θ = (1 + cos 2θ)(A.4) 2 1 sin2 θ cos2 θ = (1 − cos 4θ)(A.5) 8 1 cos θ cos ϕ = [cos(θ − ϕ) + cos(θ + ϕ)] (A.6) 2 1 sin θ sin ϕ = [cos(θ − ϕ) − cos(θ + ϕ)] (A.7) 2 D.A. Guimaraes,˜ Digital Transmission, Signals and Communication Technology, 841 DOI 10.1007/978-3-642-01359-1, C Springer-Verlag Berlin Heidelberg 2009 842 A Mathematical Tables and Algorithms 1 sin θ cos ϕ = [sin(θ + ϕ) + sin(θ − ϕ)] (A.8) 2 sin(θ ± ϕ) = sin θ cos ϕ ± cos θ sin ϕ (A.9) cos(θ ± ϕ) = cos θ cos ϕ ∓ sin θ sin ϕ (A.10) tan θ ± tan ϕ tan(θ ± ϕ) = (A.11) 1 ∓ tan θ tan ϕ sin 2θ = 2sinθ cos θ (A.12) cos 2θ = cos2 θ − sin2 θ (A.13) 1 sin θ = (eiθ − e−iθ ) (A.14) 2i 1 cos θ = (eiθ + e−iθ ).
    [Show full text]