Distributed Agreement Based Ml Approximation

سال انتشار: 1395
نوع سند: مقاله ژورنالی
زبان: انگلیسی
مشاهده: 365

فایل این مقاله در 12 صفحه با فرمت PDF قابل دریافت می باشد

استخراج به نرم افزارهای پژوهشی:

لینک ثابت به این مقاله:

شناسه ملی سند علمی:

JR_JACR-7-4_006

تاریخ نمایه سازی: 11 تیر 1396

چکیده مقاله:

In this manuscript we suggest a fast adaptive distributed method for maximumlikelihood approximation (MLA) in multiple view object localization problem. Forthis purpose, we use up to scale property of projective geometry and by definingcoefficients for convergence criterion, we increase the convergence speed of theconsensus algorithm. We try to present a mathematical model for the problem. Weuse two types of error function. The proposed method uses maximum likelihood forobtaining its best parameters. Our approach utilizes up to scale property inprojective geometry to reach the consensus quickly. The difference between nodes values and meanwhile consensus values are evaluated by two error functions. Toestimate consensus value in the second error function, we used local weightedaverage of each node. At the last of the paper, we prove our claims by experimentalresults

نویسندگان

Mohamad Mohamadi

Department of Computer Engineering, Nourabad Mamasani Branch, Islamic Azad University,Nourabad Mamasani, Iran

Hamid Parvin

Department of Computer Engineering, Nourabad Mamasani Branch, Islamic Azad University,Nourabad Mamasani, Iran Young Researchers and Elite Club, Nourabad Mamasani Branch, Islamic Azad University,Nourabad Mamasani, Iran

Es hagh Faraji

Department of Computer Engineering, Nourabad Mamasani Branch, Islamic Azad University,Nourabad Mamasani, Iran Young Researchers and Elite Club, Nourabad Mamasani Branch, Islamic Azad University,Nourabad Mamasani, Iran

Sajad Parvin

Department of Computer Engineering, Nourabad Mamasani Branch, Islamic Azad University,Nourabad Mamasani, Iran