Learning Aspect Graph Representations from View Sequences
نویسندگان
چکیده
In our effort to develop a modular neural system for invariant learning and recognition of 3D objects, we introduce here a new module architecture called an aspect network constructed around adaptive axo-axo-dendritic synapses. This builds upon our existing system (Seibert & Waxman, 1989) which processes 20 shapes and classifies t.hem into view categories (i.e ., aspects) invariant to illumination, position, orientat.ion, scale, and projective deformations. From a sequence 'of views, the aspect network learns the transitions between these aspects, crystallizing a graph-like structure from an initially amorphous network . Object recognition emerges by accumulating evidence over multiple views which activate competing object hypotheses.
منابع مشابه
Feature-based computation of hierarchical aspect-graphs
The representation of 3-D objects by characteristic or principal 2-D views needs investigations concerning applications with large object data bases with regard to the eecient access to the processed views. The reported research is directed at a context-sensitive characteristic view distribution of the data base objects in a hierarchical structure. This paper presents a new approach to the auto...
متن کاملHMMs and Vision: Representing Structure and Sequences for Active Vision Using Hidden Markov Models
Hidden Markov models (HMMs) have rarely been used in computer vision, although they are a key part of modern speech recognition systems. We show how HMMs can be used in a variety of vision related tasks. In all cases, the HMM learns structured knowledge about the world from a set of observation sequences. The HMM can also be used to generate behavior. We present three examples. • An aspect grap...
متن کاملTitle: View{based Cognitive Map Learning by an Autonomous Robot
This paper presents a view{based approach to map learning and navigation in mazes. By means of graph theory we have shown that the view{graph is a suucient representation for map behaviour such as path planning. A neural network for unsupervised learning of the view{graph from sequences of views is constructed. We use a modiied Kohonen (1988) learning rule that transforms temporal sequence (rat...
متن کاملA General Model for Multiple View Unsupervised Learning
Multiple view data, which have multiple representations from different feature spaces or graph spaces, arise in various data mining applications such as information retrieval, bioinformatics and social network analysis. Since different representations could have very different statistical properties, how to learn a consensus pattern from multiple representations is a challenging problem. In thi...
متن کاملRDF2Vec: RDF Graph Embeddings for Data Mining
Linked Open Data has been recognized as a valuable source for background information in data mining. However, most data mining tools require features in propositional form, i.e., a vector of nominal or numerical features associated with an instance, while Linked Open Data sources are graphs by nature. In this paper, we present RDF2Vec, an approach that uses language modeling approaches for unsu...
متن کامل