Compositional Hierarchical Tensor Factorization: Representing Hierarchical Intrinsic and Extrinsic Causal Factors

11/11/2019
by   Your name, et al.
1

Visual objects are composed of a recursive hierarchy of perceptual wholes and parts, whose properties, such as shape, reflectance, and color, constitute a hierarchy of intrinsic causal factors of object appearance. However, object appearance is the compositional consequence of both an object's intrinsic and extrinsic causal factors, where the extrinsic causal factors are related to illumination, and imaging conditions. Therefore, this paper proposes a unified tensor model of wholes and parts, and introduces a compositional hierarchical tensor factorization that disentangles the hierarchical causal structure of object image formation, and subsumes multilinear block tensor decomposition as a special case. The resulting object representation is an interpretable combinatorial choice of wholes' and parts' representations that renders object recognition robust to occlusion and reduces training data requirements. We demonstrate ourapproach in the context of face recognition by training on an extremely reduced dataset of synthetic images, and report encouragingface verification results on two datasets - the Freiburg dataset, andthe Labeled Face in the Wild (LFW) dataset consisting of real world images, thus, substantiating the suitability of our approach for data starved domains.

READ FULL TEXT

page 3

page 7

research
04/14/2021

Compositional Hierarchical #Tensor Factorization: Representing Hierarchical Intrinsic and Extrinsic #Causal Factors #macinelearning #womenwhocode

Visual objects are composed of a recursive hierarchy of perceptual whole...
research
02/25/2021

CausalX: Causal Explanations and Block Multilinear Factor Analysis

By adhering to the dictum, "No causation without manipulation (treatment...
research
01/01/2023

Causal Deep Learning: Causal Capsules and Tensor Transformers

We derive a set of causal deep neural networks whose architectures are a...
research
09/15/2019

A Dual-hierarchy Semantic Graph for Robust Object Recognition

We present a system for object recognition based on a semantic model gra...
research
05/23/2019

Hangul Fonts Dataset: a Hierarchical and Compositional Dataset for Interrogating Learned Representations

Interpretable representations of data are useful for testing a hypothesi...
research
01/22/2017

Greedy Compositional Clustering for Unsupervised Learning of Hierarchical Compositional Models

This paper proposes to integrate a feature pursuit learning process into...
research
01/21/2015

A Graph Theoretic Approach for Object Shape Representation in Compositional Hierarchies Using a Hybrid Generative-Descriptive Model

A graph theoretic approach is proposed for object shape representation i...

Please sign up or login with your details

Forgot password? Click here to reset