Deep Learning Based Efficient Algorithm for Object Detection Targeting Navigation Applications

Rathee, Pravesh and Acharyya, Amit (2019) Deep Learning Based Efficient Algorithm for Object Detection Targeting Navigation Applications. Masters thesis, Indian institute of technology Hyderabad.

[img] Text
Mtech_Thesis_TD1552_2019.doc
Restricted to Repository staff only until 31 July 2022.

Download (4MB) | Request a copy

Abstract

We show that deep learning is an efficient algorithm for image detection targeting navigation applications, which doesn’t have communication, not even the data transfer from external sources like any device or from the cloud. To achieve this target algorithm only needs model weights of size 105MB and can detect the image with accuracy of 72% with response time of 1.3milliseconds on GPU. At the core our approach uses convolution neural network for detection of image which is at least 10X times faster than other state of art searching algorithms, LSH or any matching algorithm. Thus this permits us to use this algorithm for image detection in navigation applications. This allows us to navigate faster, secure as well as with maintained privacy of navigator.

[error in script]
IITH Creators:
IITH CreatorsORCiD
Acharyya, Amithttp://orcid.org/0000-0002-5636-0676
Item Type: Thesis (Masters)
Uncontrolled Keywords: Deep learning, Object detection, Navigation
Subjects: Electrical Engineering
Divisions: Department of Electrical Engineering
Depositing User: Team Library
Date Deposited: 01 Aug 2019 06:59
Last Modified: 01 Aug 2019 06:59
URI: http://raiithold.iith.ac.in/id/eprint/5849
Publisher URL:
Related URLs:

    Actions (login required)

    View Item View Item
    Statistics for RAIITH ePrint 5849 Statistics for this ePrint Item