Abstract | ||
---|---|---|
Automatic feature creation is a powerful tool for identifying and reaching goals in the natural world. This paper describes in detail a biologically-inspired method of feature creation that can be applied to sensory information of any modality. The algorithm is incremental and on-line; it enforces sparseness in the features it creates; and it can form features from other features, making a hierarchical feature set. Here it demonstrates the creation of both visual and auditory features. |
Year | DOI | Venue |
---|---|---|
2011 | 10.3233/978-1-60750-959-2-305 | BIOLOGICALLY INSPIRED COGNITIVE ARCHITECTURES 2011 |
Keywords | Field | DocType |
feature creation,unsupervised learning,perception,abstraction,vision processing,sensor fusion | Cognitive science,Computer science,Perception | Conference |
Volume | ISSN | Citations |
233 | 0922-6389 | 4 |
PageRank | References | Authors |
0.50 | 5 | 1 |
Name | Order | Citations | PageRank |
---|---|---|---|
Brandon Rohrer | 1 | 60 | 8.93 |