Vector quantization with variable-precision classification.

Vector quantization with variable-precision classification. Research Abstract Details 

Research Abstract Table of Contents

Jump to the:

Vector quantization with variable-precision classification. Abstract Text:

R Dionysian,M D Ercegovac,

We investigate variable-precision classification (VPC) for speeding vector quantization (VQ). VPC evaluates bit-serially, from the most significant bit. When the magnitude of the error due to the unevaluated bits is less than the absolute magnitude of the discriminant, we can classify without processing the remaining bits. A proof shows that as the operand precision increases, the average necessary precision becomes asymptotically independent of the operand precision, VPC makes the complexity of the L(2) norm equivalent to the L(1) norm. In VQ of real images, on average, the codevector element's precision necessary for classification was under four bits. We implemented binary classification circuitry using VPC and conventional approaches. The key modules were designed and their performance estimated assuming 1.0-mum gate array technology. The implementations could search binary pruned trees at the television quality video rate. When the overall execution time is important, VPC more than halves the computational complexity.

Vector quantization with variable-precision classification. Publishing Authors By Initials

R Dionysian,MD Ercegovac,

For similar abstracts research abstracts see: abstracts research

PUBMED ID PMID:

MEDLINE DATE:

Vector quantization with variable-precision classification. Journal Published:

PUBLICATION TYPE: Journal Article

Journal: IEEE transactions on image processing : a publicat

VOLUME: 5

Page Numbers: 1528-38

Journal Abbreviation:

ISSN: 1057-7149

DAY: 21

MONTH: 02

YEAR: 1996

Vector quantization with variable-precision classification. Information

Number of References:

LANGUAGE: eng

NlmUniqueID: 9886191

Vector quantization with variable-precision classification. Keywords Mesh Terms:

KEYWORDS:

MESH TERMS:

Chemical & Substance for Abstract: Vector quantization with variable-precision classification. Information

Substance Name:

Registry Number:

Grant and Affiliation Information for Vector quantization with variable-precision classification.

AFFILIATION: Unisys Corp., Mission Viejo, CA.

Country: United States

AGENCY:

GRANT:

ACRONYM:

MEDLINETA: IEEE Trans Image Process

REFSOURCE:

DATABASENAME:

ACCESSION NUMBER:

Number Hits: 0

Vector quantization with variable-precision classification Related Publications

• A contribution to the knowledge of direct peripheral communication with the central nervous system using a strictly neurotropic rabies virus as a model.
• Further contributions to knowledge of the time during which rabies virus is bound.
• Principles of the application of antirabies serum.
• Quantitative electroencephalographic analyses in cocaine-preferring polysubstance abusers during abstinence.
• The time for which rabies virus can survive locally before its introduction into the central nervous system.
• Vector quantization with variable-precision classification.
• [Considerations and therapy of fractures of the spine with neurological manifestations (indications for laminectomy).]
• [Maxillofacial injuries associated with craniocerebral damages]
• [Neurotrauma in combination with multiple injuries]
• [Problems in the confirmation of brain death]
• [Rupture of small intestine caused by increased intra-abdominal pressure.]
• [Spontaneous atypical intracerebral hemorrhages]
• [Spontaneous atypical intracerebral hemorrhages]
• [Spontaneous atypical intracerebral hemorrhages]