IJSDR
IJSDR
INTERNATIONAL JOURNAL OF SCIENTIFIC DEVELOPMENT AND RESEARCH
International Peer Reviewed & Refereed Journals, Open Access Journal
ISSN Approved Journal No: 2455-2631 | Impact factor: 8.15 | ESTD Year: 2016
open access , Peer-reviewed, and Refereed Journals, Impact factor 8.15

Issue: March 2024

Volume 9 | Issue 3

Impact factor: 8.15

Click Here For more Info

Imp Links for Author
Imp Links for Reviewer
Research Area
Subscribe IJSDR
Visitor Counter

Copyright Infringement Claims
Indexing Partner
Published Paper Details
Paper Title: Real Time Object Detection Deep Reinforcement Learning Model: RODRLM
Authors Name: Mrs Sonal Tiwari
Unique Id: IJSDR2208070
Published In: Volume 7 Issue 8, August-2022
Abstract: Visual tracking is a challenging problem since it usually faces adverse factors, such as object deformation, fast motion, occlusion, and background clutter in practical applications. Reinforcement learning based Action-Decision Network (ADNet) has shown great potential for object tracking. However ,ADNet has some shortcomings in optimal action selection and action reward, and suffers from in efficient tracking. To this end, real time object detection deep reinforcement learning model come in tracking method , RODRLM improve efficiency and accuracy in real time object tracking . Reinforcement Learning based real time object detection framework using deep machine learning model is a unique and important technique using which user can get quality output and can use in many system for getting efficient results. In this work we propose object detection with deep reinforcement learning by which we train the agent to extract the features of sequence of the frame and with a trained agent we detect the object present in video. The proposed technique improve the feature extraction ability of its convolution layers. Then, in the reinforcement learning based training phase, both the selection criteria for optimal action and the reward function are redesigned separately to explore more appropriate action and eliminate useless action .Finally, an effective online adaptive update strategy is proposed to adapt to the appearance changes or deformation of the object during actual tracking. Specially, meta-learning is utilized to pursue the most appropriate parameters for the network so that the parameters are closer to the optimal ones in the subsequent tracking process.
Keywords: Visual tracking, reinforcement learning, meta-learning, multi-domain training.
Cite Article: "Real Time Object Detection Deep Reinforcement Learning Model: RODRLM", International Journal of Science & Engineering Development Research (www.ijsdr.org), ISSN:2455-2631, Vol.7, Issue 8, page no.515 - 521, August-2022, Available :http://www.ijsdr.org/papers/IJSDR2208070.pdf
Downloads: 000336262
Publication Details: Published Paper ID: IJSDR2208070
Registration ID:201171
Published In: Volume 7 Issue 8, August-2022
DOI (Digital Object Identifier):
Page No: 515 - 521
Publisher: IJSDR | www.ijsdr.org
ISSN Number: 2455-2631

Click Here to Download This Article

Article Preview

Click here for Article Preview







Major Indexing from www.ijsdr.org
Google Scholar ResearcherID Thomson Reuters Mendeley : reference manager Academia.edu
arXiv.org : cornell university library Research Gate CiteSeerX DOAJ : Directory of Open Access Journals
DRJI Index Copernicus International Scribd DocStoc

Track Paper
Important Links
Conference Proposal
ISSN
DOI (A digital object identifier)


Providing A digital object identifier by DOI
How to GET DOI and Hard Copy Related
Open Access License Policy
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License
Creative Commons License
This material is Open Knowledge
This material is Open Data
This material is Open Content
Social Media
IJSDR

Indexing Partner