DOCSLIB.ORG

PHY–396 K. Solutions for Homework Set #6. Problem 1(A): [ ˆ J , ˆ J ≡ 1 Ε

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
PHY–396 K. Solutions for Homework Set #6. Problem 1(A): [ ˆ J , ˆ J ≡ 1 Ε PHY{396 K. Solutions for homework set #6. Problem 1(a): h ^i ^ji 1 ik` jmn h ^k` ^mni J ; J ≡ 4 J ; J = hh by eq. (1) ii 1 ik` jmn km ^`n kn ^`m `m ^kn `n ^km = 4 −ig J + ig J + ig J − g J hh by antisymmetry of the 's ii = ik`jmn × −igkmJ^`n = iJ^`n × −gkmik`jmn = +δkmik`jmn = δijδ`n − δinδ`j = 0 − iJ^ji = +iJ^ij ≡ +iijkJ^k: (S:1) h ^i ^ ji 1 ik` h ^k` ^0ji J ; K ≡ 2 J ; J = hh by eq. (1) ii 1 ik` k0 ^`j kj ^`0 `0 ^kj `j ^k0 = 2 −ig J + ig J + ig J − ig J 1 ik` kj ^`0 `j ^k0 = 2 0 − iδ J + 0 + iδ J 1 ik` kj ^ ` `j ^ k ≡ 2 +iδ K − iδ K 1 ij` ^ ` 1 ikj ^ j = 2 K − 2 K = ijkK^ k; (S:2) h i h i K^ i; K^ j ≡ J^0i; J^0j = hh by eq. (1) ii = −ig00J^ij + ig0jJ^i0 + igi0J^0j − igijJ^00 = −iJ^ij + 0 + 0 + 0; ≡ −iijkJ^k: (S:3) Problem 1(b): h ^ i ^ji 1 jk` h ^ i ^k`i 1 jk` ik ^ ` i` ^ k V ; J ≡ 2 V ; J = 2 ig V − ig V 1 jk` ik ^ ` i` ^ k i ji` ^ ` i jki ^ k = 2 −iδ V + δ V = − 2 V + 2 V = iijkV^ k; (S:4) 1 h ^ 0 ^ji 1 jk` h ^ 0 ^k`i 1 jk` 0k ^ ` 0` ^ k V ; J = 2 V ; J = 2 i\g V − i\g V = 0; (S:5) h i h i V^ i; K^ j = V^ i; J^0j = i\gi0V^ j − igijV^ 0 = +iδijV^ 0; (S:6) h i h i V^ 0; K^ j = V^ 0; J^0j = ig00V^ j − i\g0jV^ 0 = +iV^ j: (S:7) Note that the Hamiltonian of a relativistic theory is a member of a 4-vector multiplet P^µ = (H;^ P^) where P^ is the net momentum operator. Applying the above equations to the P^µ vectors, we obtain h i P^i; J^j = iijkP^k; h i H;^ J^j = 0; h i (S:8) P^i; K^ j = +iδijH;^ h i H;^ K^ j = +iP^j: In particular, the Hamiltonian H^ commutes with the three angular momenta J^j but it does not commute with the three generators K^ k of the Lorentz boosts. Problem 1(c): In the ordinary quantum mechanics, it is often said that generators of continuous symmetries must commute with the Hamiltonian operator. However, this is true only for the symmetries that act in a time independent manner | for example, rotating the 3D space by the same angle at all times t. But when the transformation rules of a symmetry depend on time, the Hamiltonian must change to account for this time dependence. In a Lorentz boost, the transform x ! x0 obviously depends on time, which changes the way the transformed quantum fields such as Φ^ 0(x; t) depend on t. Consequently, the Hamiltonian H^ of the theory must change so that the new Heisenberg equations would match the new time dependence. In terms of the generators, this means that the boost generators K^ i should not commute with the Hamiltonian. 2 Note that this non-commutativity is not caused by the Lorentz boosts affecting the time 0 0 µ itself, t = L µx 6= t. Even in non-relativistic theories | where the time is absolute | the generators of symmetries which affect the other variables in a time-dependent matter do not commute with the Hamiltonian H^ . Indeed, consider a Galilean transform from one non-relativistic moving frame into an- other, x0 = x+vt but t0 = t. This is a good symmetry of non-relativistic particles interacting with each other but not subject to any external potential, X 1 X H^ = ^p2 + 1 V (x^ − x^ ): (S:9) 2M a 2 a b a a6=b A unitary operator G^ realizing a Galilean symmetry acts on coordinate and momentum operators as y y G^x^aG^ = x^a + vt; G^^paG^ = ^pa + Mv; (S:10) and it also transforms the Hamiltonian into ^ ^ ^y ^ ^ 1 2 GHG = H + v · Ptot + 2 Mtotv : (S:11) In terms of the Galilean boost generators K^ G, G^ = exp −iv · K^ G ; (S:12) hence under an infinitesimal boost v = ~, various operators transform according to i h ^ i i O!b Ob + δOb; δOb = −i KG; Ob : (S:13) Hence, the commutation relations with the boost generators follow from the infinitesimal boosts, for example h i ^ j i ij δx^a = ~t =) x^a; KG = iδ × t; h i ^ j i ij δp^a = ~M =) p^a; KG = iδ × M; (S:14) ^ ^ h ^ ^ j i ^i δH = ~ · Ptot =) H; KG = iPtot : In particular, the Hamiltonian does NOT commute with the Galilean boost generators. 3 Problem 2(a): 1 ^µν Consider a linear combination 2 NµνJ of Lorentz generators with some generic coefficients iφ ^µν Nµν = −Nνµ. The Lorentz symmetries L(N; φ) = exp( 2 NµνJ ) generated by this combi- nation preserve the momentum pµ of the particle state jpi if and only if iφ ^µν ^α iφ ^µν ^µ exp 2 NµνJ P exp − 2 NµνJ jpi = P jpi : (S:15) For an infinitesimal φ, this condition becomes h iφ ^µν ^αi 2 NµνJ ; P jpi = 0: (S:16) Applying eqs. (13) to the left hand side of this formula gives us h iφ ^µν ^αi iφ αµ ^ν αν ^µ 2 NµνJ ; P jpi = 2 Nµν −ig P + g P jpi αν = φN P^ν jpi (S:17) αν = φN pν jpi ; 1 ^µν µ so the condition for the generator 2 NµνJ to preserve particle's momentum p is simply αν N × pν = 0: (S:18) In 3D terms, N ij = ijkak and N 0k = −N k0 = bk for some 3-vectors a and b, the generator in question is 1 ^µν ^ ^ 2 NµνJ = a · J + b · K; (S:19) and the condition (S.18) becomes a × p − bE = 0 and b · p = 0: (S:20) Actually, the second condition here is redundant, so the general solution is p any a; b = a × : (S:21) E In terms of eq. (S.19), these solutions mean (a × p) · K^ a 1 N J^µν = a · J^ + = · EJ^ + p × K^ for any a: (S:22) 2 µν E E In other words, the Lorentz symmetries preserving the momentum pµ have 3 generators, 4 namely the components of the 3-vector R^ = E J^ + p × K^ : (4) Quod erat demonstrandum. Problem 2(b): Consider a massive particle moving at a slower-than-light speed β in z direction, so the energy is E = γm (where γ = 1=p1 − β2) while the 3-momentum is p = (0; 0; βγm). Consequently, the three components of the R^ vector this energy-momentum are R^x = γmJ^x − βγmK^ y; R^y = γmJ^y + βγmK^ x; R^z = γmJ^z: (S:23) These 3 operators generate the little group G(p) of the particle's 4{momentum pµ = (E; p). To see that this little group happens to be isomorphic to the 3D rotation group SO(3), we need to find 3 linear combinations J~x;y;z of the operators (S.23) which obey the angular- momentum commutation relations h i J~i; J~j = iijkJ~k: (S:24) My choice of the (properly normalized) generators J~x;y;z is spelled out in eqs. (5). Let's see that they indeed obey the commutation relations (S.24): h i h i h i J^z; Jex = γ J^z; J^x − βγ J^z; K^ y = γ × iJ^y − βγ × (−iK^ x) = iJey; h i h i h i J^z; Jey = γ J^z; J^y + βγ J^z; K^ x = γ × (−iJ^x) − βγ × (+iK^ y) (S:25) = −iJex; h x yi 2 h ^x ^yi 2 h ^ y ^yi 2 h ^ ^ x i 2 2 h ^ y ^ xi Je ; Je = γ J ; J − βγ K ; J + βγ Jx; K ; − β γ K ; K = γ2 × iJ^z − 0 + 0 − β2γ2 × iJ^z = iJ^z × γ2(1 − β2) = 1 = iJ^z: Quod erat demonstrandum. 5 Problem 2(c): For a massless particle, we cannot rescale the little group generators as in eqs. (5) since the 1=m factor for the J~x;y generators would be infinite. Instead, the best we can do is to use the 1=E factor for all three rescaled generators, hence eqs. (6) and (7). Consequently, instead of eqs. (S.25) for the commutators of the rescaled generators, we get h i h i h i J^z; I^x = J^z; J^x − J^z; K^ y = iJ^y − (−iK^ x) = iI^y; h i h i h i J^z; I^y = J^z; J^y + J^z; K^ x = (−iJ^x) − (+iK^ y) = −iI^x; (S:26) h i h i h i h i h i I^x; I^y = J^x; J^y − K^ y; J^y + J^x; K^ x − K^ y; K^ x = iJ^z − 0 + 0 − iJ^z = 0; precisely as in eq. (8). The commutation relations (8) are different from the angular-momentum commutation relations, and they cannot be brought to the form (S.24) by any finite rescaling of the generators. Consequently, the little group of a light-like momentum pµ = (E; 0; 0;E) is NOT isomorphic to the SO(3). Instead, the commutation relations (8) are similar to the commutation relations between the z component of the angular momentum and the x and y components of the linear momentum, h i h i h i J^z; P^x = +iP^y; J^z; P^y = −iP^x; P^x; P^y = 0: (S:27) The J^z operator generates rotations around the z axis, i.e., within the xy plane, while the P^x and P^i operators generate translations in that plane.
Load more

Recommended publications
  • Bearing Rigidity Theory in SE(3) Giulia Michieletto, Angelo Cenedese, Antonio Franchi
    Bearing Rigidity Theory in SE(3) Giulia Michieletto, Angelo Cenedese, Antonio Franchi To cite this version: Giulia Michieletto, Angelo Cenedese, Antonio Franchi. Bearing Rigidity Theory in SE(3). 55th IEEE Conference on Decision and Control, Dec 2016, Las Vegas, United States. hal-01371084 HAL Id: hal-01371084 https://hal.archives-ouvertes.fr/hal-01371084 Submitted on 23 Sep 2016 HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Preprint version, final version at http://ieeexplore.ieee.org/ 55th IEEE Conference on Decision and Control. Las Vegas, NV, 2016 Bearing Rigidity Theory in SE(3) Giulia Michieletto, Angelo Cenedese, and Antonio Franchi Abstract— Rigidity theory has recently emerged as an ef- determines the infinitesimal rigidity properties of the system, ficient tool in the control field of coordinated multi-agent providing a necessary and sufficient condition. In such a systems, such as multi-robot formations and UAVs swarms, that context, a framework is generally represented by means of are characterized by sensing, communication and movement capabilities. This work aims at describing the rigidity properties the bar-and-joint model where agents are represented as for frameworks embedded in the three-dimensional Special points joined by bars whose fixed lengths enforce the inter- Euclidean space SE(3) wherein each agent has 6DoF.
    [Show full text]
  • Chapter 5 ANGULAR MOMENTUM and ROTATIONS
    Chapter 5 ANGULAR MOMENTUM AND ROTATIONS In classical mechanics the total angular momentum L~ of an isolated system about any …xed point is conserved. The existence of a conserved vector L~ associated with such a system is itself a consequence of the fact that the associated Hamiltonian (or Lagrangian) is invariant under rotations, i.e., if the coordinates and momenta of the entire system are rotated “rigidly” about some point, the energy of the system is unchanged and, more importantly, is the same function of the dynamical variables as it was before the rotation. Such a circumstance would not apply, e.g., to a system lying in an externally imposed gravitational …eld pointing in some speci…c direction. Thus, the invariance of an isolated system under rotations ultimately arises from the fact that, in the absence of external …elds of this sort, space is isotropic; it behaves the same way in all directions. Not surprisingly, therefore, in quantum mechanics the individual Cartesian com- ponents Li of the total angular momentum operator L~ of an isolated system are also constants of the motion. The di¤erent components of L~ are not, however, compatible quantum observables. Indeed, as we will see the operators representing the components of angular momentum along di¤erent directions do not generally commute with one an- other. Thus, the vector operator L~ is not, strictly speaking, an observable, since it does not have a complete basis of eigenstates (which would have to be simultaneous eigenstates of all of its non-commuting components). This lack of commutivity often seems, at …rst encounter, as somewhat of a nuisance but, in fact, it intimately re‡ects the underlying structure of the three dimensional space in which we are immersed, and has its source in the fact that rotations in three dimensions about di¤erent axes do not commute with one another.
    [Show full text]
  • For Deep Rotation Learning with Uncertainty
    A Smooth Representation of Belief over SO(3) for Deep Rotation Learning with Uncertainty Valentin Peretroukhin,1;3 Matthew Giamou,1 David M. Rosen,2 W. Nicholas Greene,3 Nicholas Roy,3 and Jonathan Kelly1 1Institute for Aerospace Studies, University of Toronto; 2Laboratory for Information and Decision Systems, 3Computer Science & Artificial Intelligence Laboratory, Massachusetts Institute of Technology Abstract—Accurate rotation estimation is at the heart of unit robot perception tasks such as visual odometry and object pose quaternions ✓ ✓ q = q<latexit sha1_base64="BMaQsvdb47NYnqSxvFbzzwnHZhA=">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</latexit> ⇤ ✓ estimation. Deep neural networks have provided a new way to <latexit sha1_base64="1dHo/MB41p5RxdHfWHI4xsZNFIs=">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</latexit>
    [Show full text]
  • Introduction to SU(N) Group Theory in the Context of The
    215c, 1/30/21 Lecture outline. c Kenneth Intriligator 2021. ? Week 1 recommended reading: Schwartz sections 25.1, 28.2.2. 28.2.3. You can find more about group theory (and some of my notation) also here https://keni.ucsd.edu/s10/ • At the end of 215b, I discussed spontaneous breaking of global symmetries. Some students might not have taken 215b last quarter, so I will briefly review some of that material. Also, this is an opportunity to introduce or review some group theory. • Physics students first learn about continuous, non-Abelian Lie groups in the context of the 3d rotation group. Let ~v and ~w be N-component vectors (we'll set N = 3 soon). We can rotate the vectors, or leave the vectors alone and rotate our coordinate system backwards { these are active vs passive equivalent perspectives { and scalar quantities like ~v · ~w are invariant. This is because the inner product is via δij and the rotation is a similarity transformation that preserves this. Any rotation matrix R is an orthogonal N × N matrix, RT = R−1, and the space of all such matrices is the O(N) group manifold. Note that det R = ±1 has two components, and we can restrict to the component with det R = 1, which is connected to the identity and called SO(N). The case SO(2), rotations in a plane, is parameterized by an angle θ =∼ θ + 2π, so the group manifold is a circle. If we use z = x + iy, we see that SO(2) =∼ U(1), the group of unitary 1 × 1 matrices eiθ.
    [Show full text]
  • Quaternion Product Units for Deep Learning on 3D Rotation Groups
    Quaternion Product Units for Deep Learning on 3D Rotation Groups Xuan Zhang1, Shaofei Qin1, Yi Xu∗1, and Hongteng Xu2 1MoE Key Lab of Artificial Intelligence, AI Institute 2Infinia ML, Inc. 1Shanghai Jiao Tong University, Shanghai, China 2Duke University, Durham, NC, USA ffloatlazer, 1105042987, [email protected], [email protected] Abstract methods have made efforts to enhance their robustness to rotations [4,6, 29,3, 26]. However, the architectures of We propose a novel quaternion product unit (QPU) to their models are tailored for the data in the Euclidean space, represent data on 3D rotation groups. The QPU lever- rather than 3D rotation data. In particular, the 3D rotation ages quaternion algebra and the law of 3D rotation group, data are not closed under the algebraic operations used in representing 3D rotation data as quaternions and merg- their models (i:e:, additions and multiplications).1 The mis- ing them via a weighted chain of Hamilton products. We matching between data and model makes these methods dif- prove that the representations derived by the proposed QPU ficult to analyze the influence of rotations on their outputs can be disentangled into “rotation-invariant” features and quantitatively. Although this mismatching problem can be “rotation-equivariant” features, respectively, which sup- mitigated by augmenting training data with additional rota- ports the rationality and the efficiency of the QPU in the- tions [18], this solution gives us no theoretical guarantee. ory. We design quaternion neural networks based on our To overcome the challenge mentioned above, we pro- QPUs and make our models compatible with existing deep posed a novel quaternion product unit (QPU) for 3D ro- learning models.
    [Show full text]
  • Auto-Encoding Transformations in Reparameterized Lie Groups for Unsupervised Learning
    The Thirty-Fifth AAAI Conference on Artificial Intelligence (AAAI-21) Auto-Encoding Transformations in Reparameterized Lie Groups for Unsupervised Learning Feng Lin,1 Haohang Xu,2 Houqiang Li,1,4 Hongkai Xiong,2 Guo-Jun Qi3,* 1 CAS Key Laboratory of GIPAS, EEIS Department, University of Science and Technology of China 2 Department of Electronic Engineering, Shanghai Jiao Tong University 3 Laboratory for MAPLE, Futurewei Technologies 4 Institute of Artificial Intelligence, Hefei Comprehensive National Science Center Abstract Unsupervised training of deep representations has demon- strated remarkable potentials in mitigating the prohibitive ex- penses on annotating labeled data recently. Among them is predicting transformations as a pretext task to self-train repre- sentations, which has shown great potentials for unsupervised learning. However, existing approaches in this category learn representations by either treating a discrete set of transforma- tions as separate classes, or using the Euclidean distance as the metric to minimize the errors between transformations. None of them has been dedicated to revealing the vital role of the geometry of transformation groups in learning represen- Figure 1: The deviation between two transformations should tations. Indeed, an image must continuously transform along be measured along the curved manifold (Lie group) of the curved manifold of a transformation group rather than through a straight line in the forbidden ambient Euclidean transformations rather than through the forbidden Euclidean space. This suggests the use of geodesic distance to minimize space of transformations. the errors between the estimated and groundtruth transfor- mations. Particularly, we focus on homographies, a general group of planar transformations containing the Euclidean, works to explore the possibility of pretraining representa- similarity and affine transformations as its special cases.
    [Show full text]
  • Finite Subgroups of the Group of 3D Rotations
    FINITE SUBGROUPS OF THE 3D ROTATION GROUP Student : Nathan Hayes Mentor : Jacky Chong SYMMETRIES OF THE SPHERE • Let’s think about the rigid rotations of the sphere. • Rotations can be composed together to form new rotations. • Every rotation has an inverse rotation. • There is a “non-rotation,” or an identity rotation. • For us, the distinct rotational symmetries of the sphere form a group, where each rotation is an element and our operation is composing our rotations. • This group happens to be infinite, since you can continuously rotate the sphere. • What if we wanted to study a smaller collection of symmetries, or in some sense a subgroup of our rotations? THEOREM • A finite subgroup of SO3 (the group of special rotations in 3 dimensions, or rotations in 3D space) is isomorphic to either a cyclic group, a dihedral group, or a rotational symmetry group of one of the platonic solids. • These can be represented by the following solids : Cyclic Dihedral Tetrahedron Cube Dodecahedron • The rotational symmetries of the cube and octahedron are the same, as are those of the dodecahedron and icosahedron. GROUP ACTIONS • The interesting portion of study in group theory is not the study of groups, but the study of how they act on things. • A group action is a form of mapping, where every element of a group G represents some permutation of a set X. • For example, the group of permutations of the integers 1,2, 3 acting on the numbers 1, 2, 3, 4. So for example, the permutation (1,2,3) -> (3,2,1) will swap 1 and 3.
    [Show full text]
  • 3D Rotations Matrices
    Chapter 3 3D Rotations matrices The goal of this chapter is to exemplify on a simple and useful manifold part of the general meth- ods developed previously on Riemannian geometry and Lie groups. This document also provide numerically stable formulas to implement all the basic operations needed to work with rotations in 3D. 3.1 Vector rotations of R3 3.1.1 Definition 3 Let fi; j; kg be a right handed orthonormal basis of the vector space R (we do not consider it here as a Euclidean space), and B = fe1; e2; e3g be a set of three vectors. The linear mapping from fi; j; kg to B is given by the matrix 2 3 e11 e21 e31 R = [e1; e2; e3] = 4 e12 e22 e32 5 (3.1) e13 e23 e33 If we now want B to be an orthonormal basis, we have that h ei j ej i = δij, which means that the matrix R verifies T T R:R = R :R = I3 (3.2) This equation implies in turn that det(R:RT) = 1 (i.e. det(R) = ±1) and gives rise to the subset O3 of orthogonal matrices of M3×3. If we want B to be also right handed, we have to impose det(R) = 1. T SO3 = fR 2 M3×3=R:R = Id and det(R) = +1g (3.3) Such matrices are called (proper) rotations. Each right handed orthonormal basis can then be represented by a unique rotation and conversely each rotation maps fi; j; kg onto a unique right handed orthonormal basis.
    [Show full text]
  • Mathematical Basics of Motion and Deformation in Computer Graphics Second Edition
    Mathematical Basics of Motion and Deformation in Computer Graphics Second Edition Synthesis Lectures on Visual Computing Computer Graphics, Animation, Computational Photography, and Imaging Editor Brian A. Barsky, University of California, Berkeley is series presents lectures on research and development in visual computing for an audience of professional developers, researchers, and advanced students. Topics of interest include computational photography, animation, visualization, special effects, game design, image techniques, computational geometry, modeling, rendering, and others of interest to the visual computing system developer or researcher. Mathematical Basics of Motion and Deformation in Computer Graphics: Second Edition Ken Anjyo and Hiroyuki Ochiai 2017 Digital Heritage Reconstruction Using Super-resolution and Inpainting Milind G. Padalkar, Manjunath V. Joshi, and Nilay L. Khatri 2016 Geometric Continuity of Curves and Surfaces Przemyslaw Kiciak 2016 Heterogeneous Spatial Data: Fusion, Modeling, and Analysis for GIS Applications Giuseppe Patanè and Michela Spagnuolo 2016 Geometric and Discrete Path Planning for Interactive Virtual Worlds Marcelo Kallmann and Mubbasir Kapadia 2016 An Introduction to Verification of Visualization Techniques Tiago Etiene, Robert M. Kirby, and Cláudio T. Silva 2015 iv Virtual Crowds: Steps Toward Behavioral Realism Mubbasir Kapadia, Nuria Pelechano, Jan Allbeck, and Norm Badler 2015 Finite Element Method Simulation of 3D Deformable Solids Eftychios Sifakis and Jernej Barbic 2015 Efficient Quadrature
    [Show full text]
  • The Geometry of Eye Rotations and Listing's Law
    The Geometry of Eye Rotations and Listing's Law Amir A. Handzel* Tamar Flasht Department of Applied Mathematics and Computer Science Weizmann Institute of Science Rehovot, 76100 Israel Abstract We analyse the geometry of eye rotations, and in particular saccades, using basic Lie group theory and differential geome­ try. Various parameterizations of rotations are related through a unifying mathematical treatment, and transformations between co-ordinate systems are computed using the Campbell-Baker­ Hausdorff formula. Next, we describe Listing's law by means of the Lie algebra so(3). This enables us to demonstrate a direct connection to Donders' law, by showing that eye orientations are restricted to the quotient space 80(3)/80(2). The latter is equiv­ alent to the sphere S2, which is exactly the space of gaze directions. Our analysis provides a mathematical framework for studying the oculomotor system and could also be extended to investigate the geometry of mUlti-joint arm movements. 1 INTRODUCTION 1.1 SACCADES AND LISTING'S LAW Saccades are fast eye movements, bringing objects of interest into the center of the visual field. It is known that eye positions are restricted to a subset of those which are anatomically possible, both during saccades and fixation (Tweed & Vilis, 1990). According to Donders' law, the eye's gaze direction determines its orientation uniquely, and moreover, the orientation does not depend on the history of eye motion which has led to the given gaze direction. A precise specification of the "allowed" subspace of position is given by Listing's law: the observed orientations of the eye are those which can be reached from the distinguished orientation called primary *[email protected] t [email protected] 118 A.
    [Show full text]
  • On Quaternions and Octonions: Their Geometry, Arithmetic, and Symmetry,By John H
    BULLETIN (New Series) OF THE AMERICAN MATHEMATICAL SOCIETY Volume 42, Number 2, Pages 229–243 S 0273-0979(05)01043-8 Article electronically published on January 26, 2005 On quaternions and octonions: Their geometry, arithmetic, and symmetry,by John H. Conway and Derek A. Smith, A K Peters, Ltd., Natick, MA, 2003, xii + 159 pp., $29.00, ISBN 1-56881-134-9 Conway and Smith’s book is a wonderful introduction to the normed division algebras: the real numbers (R), the complex numbers (C), the quaternions (H), and the octonions (O). The first two are well-known to every mathematician. In contrast, the quaternions and especially the octonions are sadly neglected, so the authors rightly concentrate on these. They develop these number systems from scratch, explore their connections to geometry, and even study number theory in quaternionic and octonionic versions of the integers. Conway and Smith warm up by studying two famous subrings of C: the Gaussian integers and Eisenstein integers. The Gaussian integers are the complex numbers x + iy for which x and y are integers. They form a square lattice: i 01 Any Gaussian integer can be uniquely factored into ‘prime’ Gaussian integers — at least if we count differently ordered factorizations as the same and ignore the ambiguity introduced by the invertible Gaussian integers, namely ±1and±i.To show this, we can use a straightforward generalization of Euclid’s proof for ordinary integers. The key step is to show that given nonzero Gaussian integers a and b,we can always write a = nb + r for some Gaussian integers n and r, where the ‘remainder’ r has |r| < |b|.
    [Show full text]
  • Arxiv:1909.12057V4 [Cs.LG] 22 Mar 2021 Behavior Under Such Transformations and Are Insensitive to Both Local and Global Transformations on the Input Data
    Published as a conference paper at ICLR 2020 B-SPLINE CNNS ON LIE GROUPS Erik J. Bekkers Centre for Analysis and Scientific Computing Department of Applied Mathematics and Computer Science Eindhoven University of Technology, Eindhoven, the Netherlands [email protected] ABSTRACT Group convolutional neural networks (G-CNNs) can be used to improve classi- cal CNNs by equipping them with the geometric structure of groups. Central in the success of G-CNNs is the lifting of feature maps to higher dimensional disentangled representations in which data characteristics are effectively learned, geometric data-augmentations are made obsolete, and predictable behavior under geometric transformations (equivariance) is guaranteed via group theory. Cur- rently, however, the practical implementations of G-CNNs are limited to either discrete groups (that leave the grid intact) or continuous compact groups such as rotations (that enable the use of Fourier theory). In this paper we lift these lim- itations and propose a modular framework for the design and implementation of G-CNNs for arbitrary Lie groups. In our approach the differential structure of Lie groups is used to expand convolution kernels in a generic basis of B-splines that is defined on the Lie algebra. This leads to a flexible framework that enables localized, atrous, and deformable convolutions in G-CNNs by means of respec- tively localized, sparse and non-uniform B-spline expansions. The impact and potential of our approach is studied on two benchmark datasets: cancer detection in histopathology slides in which rotation equivariance plays a key role and facial landmark localization in which scale equivariance is important.
    [Show full text]