Geometric deep learning is a relatively nascent field that has attracted significant attention in the recent past. This is partly due to the ready availability of manifold-valued data. In this paper we present a novel theoretical framework for developing deep neural networks to cope with manifold-valued fields as inputs. We also present a novel architecture to realize this theory and call it a ManifoldNet. Analogous to convolutions in vector spaces which are equivalent to computing weighted sums, manifold-valued data ‘convolutions’ can be defined using the weighted Frechet Mean (wFM), an intrinsic operation. The hidden layers of ManifoldNet compute wFM of their inputs, where the weights are to be learnt. Since wFM is an intrinsic operation, the processed data remain on the manifold as they propagate through the hidden layers. To reduce the computational burden, we present a provably convergent recursive algorithm for wFM computation. Additionaly, for non-constant curvature manifolds, we prove that each wFM layer is non-collapsible and a contraction mapping. We also prove that the wFM is equivariant to the symmetry group action admitted by the manifold on which the data reside. To showcase the performance of ManifoldNet, we present several experiments from vision and medical imaging.
10 November 2020 in IEEE Transactions on Neural Networks and Learning Systems
Complex-valued data are ubiquitous in signal and image processing applications, and complex-valued representations in deep learning have appealing theoretical properties. While these aspects have long been recognized, complex-valued deep learning continues to lag far behind its real-valued counterpart. We propose a principled geometric approach to complex-valued deep learning. Complex-valued data could often be subject to arbitrary complex-valued scaling; as a result, real and imaginary components could covary. Instead of treating complex values as two independent channels of real values, we recognize their underlying geometry: we model the space of complex numbers as a product manifold of nonzero scaling and planar rotations. Arbitrary complex-valued scaling naturally becomes a group of transitive actions on this manifold. We propose to extend the property instead of the form of real-valued functions to the complex domain. We define convolution as the weighted Fréchet mean on the manifold that is equivariant to the group of scaling/rotation actions and define distance transform on the manifold that is invariant to the action group. The manifold perspective also allows us to define nonlinear activation functions, such as tangent ReLU and G-transport, as well as residual connections on the manifold-valued data. We dub our model SurReal, as our experiments on MSTAR and RadioML deliver high performance with only a fractional size of real- and complex-valued baseline models.
Convolutional neural networks have been extremely successful in image-based learning tasks due to their translation equivariance property. Recent work has generalized the traditional convolutional layer of a convolutional neural network to non-Euclidean spaces and shown group equivariance of the generalized convolution operation. In this paper, we present a novel higher order Volterra convolutional neural network (VolterraNet) for data defined as samples of functions on Riemannian homogeneous spaces. Analagous to the result for traditional convolutions, we prove that the Volterra functional convolutions are equivariant to the action of the isometry group admitted by the Riemannian homogeneous spaces and, under some restrictions, any non-linear equivariant function can be expressed as our homogeneous space Volterra convolution, generalizing the non-linear shift equivariant characterization of Volterra expansions in Euclidean space. We also prove that second order functional convolution operations can be represented as cascaded convolutions which leads to an efficient implementation. Beyond this, we propose a dilated VolterraNet model. These advances lead to large parameter reductions relative to baseline non-Euclidean CNNs. To demonstrate the efficacy of the VolterraNet performance, we present several real data experiments involving classification and group testing of spherical-MNIST, atomic energy, Shrec17, and diffusion MRI data. Performance comparisons to the state-of-the-art are presented.
4 May 2020 in IEEE Transactions on Pattern Analysis and Machine Intelligence
Principal component analysis (PCA) and Kernel principal component analysis (KPCA) are fundamental methods in machine learning for dimensionality reduction. The former is a technique for finding this approximation in finite dimensions and the latter is often in an infinite dimensional reproducing Kernel Hilbert-space (RKHS). In this paper, we present a geometric framework for computing the principal linear subspaces in both (finite and infinite) situations as well as for the robust PCA case, that amounts to computing the intrinsic average on the space of all subspaces: the Grassmann manifold. Points on this manifold are defined as the subspaces spanned by $K$ -tuples of observations. The intrinsic Grassmann average of these subspaces are shown to coincide with the principal components of the observations when they are drawn from a Gaussian distribution. We show similar results in the RKHS case and provide an efficient algorithm for computing the projection onto the this average subspace. The result is a method akin to KPCA which is substantially faster. Further, we present a novel online version of the KPCA using our geometric framework. Competitive performance of all our algorithms are demonstrated on a variety of real and synthetic data sets.
31 July 2014 in International Journal of Neural Systems