A
Avinash C. Kak
Researcher at Purdue University
Publications - 259
Citations - 26030
Avinash C. Kak is an academic researcher from Purdue University. The author has contributed to research in topics: Mobile robot & Video tracking. The author has an hindex of 51, co-authored 254 publications receiving 25027 citations. Previous affiliations of Avinash C. Kak include Infosys.
Papers
More filters
Journal ArticleDOI
A Generic Road-Following Framework for Detecting Markings and Objects in Satellite Imagery
TL;DR: This paper presents a framework for detecting and labeling objects in satellite imagery, characterized by low-level features that exist mostly at or beyond the limits of spatial resolution in the satellite images.
Patent
Process and system for assessing modularity of an object-oriented program
TL;DR: In this article, the authors describe a process, system and computer program product for assessing the modularity of an object-oriented program, which includes calculation of metrics associated with various properties of the program.
Journal ArticleDOI
A predictive duty cycle adaptation framework using augmented sensing for wireless camera networks
TL;DR: This work presents a predictive framework to provide nodes with an ability to anticipate the arrival of objects in the field-of-view of their cameras by using an existing MAC header bit that is already in the 802.15.4 protocol.
Posted Content
Homography Estimation with Convolutional Neural Networks Under Conditions of Variance.
David Niblick,Avinash C. Kak +1 more
TL;DR: This report analyzes the performance of two recently published methods using Convolutional Neural Networks that are meant to replace the more traditional feature-matching based approaches to the estimation of homography and finds that CNNs can be trained to be more robust against noise, but at a small cost to accuracy in the noiseless case.
Proceedings ArticleDOI
An approach-path independent framework for place recognition and mobile robot localization in interior hallways
TL;DR: This work provides a fast approach-path-independent framework for the problem of place recognition and robot localization in indoor environments by using highly viewpoint-invariant 3D junction features extracted from stereo pairs of images based on stereo reconstructions of the JUDOCA junctions extracted from the individual images of a stereo pair.