Computer Vision: Visual Extent of an Object
Journal Title: IOSR Journals (IOSR Journal of Computer Engineering) - Year 2013, Vol 14, Issue 4
Abstract
The visual extent of an object reaches beyond the object itself. It is reflected in image retrieval techniques which combine statistics from the whole image in order to identify the image within. Nevertheless, it is still unclear to what degree and how this visual extent of an object affects the classification performance.Here we analyze the visual extent of an object on the Pascal VOC dataset using bag of words implementation with SIFT Descriptors. Our analysis is performed from two angles: (a) Not knowing the object location, we determine where in the image the support for object classification resides (normal situation) and (b) Assuming that the object location is known, we evaluate he relative potential of the object and its surround, and of the object border and object interior (ideal situation).
Authors and Affiliations
Akshit Chopra
Keywords
Related Articles
- EP ID EP88015
- DOI -
- Views 124
- Downloads 0
Outage Probability Performance Analysis of a New Hybrid Relay Selection Protocol
In this paper, the diversity gains of wireless relay networks were exploited. The proposed cooperative scheme using the first best relay for providing the highest SNR at the destination, whereas the second best relay is...
Numerical Solution of unsteady gravity flow of power-law fluids through a porous medium
Abstract: In this paper, we present a numerical study of the rheological behaviour effect of non-Newtonian, power-law fluids on the unsteady gravity flow through a porous medium. The governing equations are derived&...
Detecting Malicious Nodes in Wireless Sensor Networks
Abstract:A typical wireless sensor network consists of several tiny and low-power sensors which use radio frequencies to perform distributed sensing tasks. Wireless sensor networks are used to detect the occurrence of ev...
Empirical Study on Classification Algorithm For Evaluation of Students Academic Performance
Abstract: Data mining techniques (DMT) are extensively used in educational field to find new hidden patterns from student’s data. In recent years, the greatest issues that educational institutions are facing the unstable...
A Name Entity Detection and Relation Extraction fromUnstructured Data by N-gram Features
Abstract : In recent years Name entity extraction and linking have received much attention. However, correctclassification of entities and proper linking among these entities is a major challenge for researcher. Wepropos...