SIGNAL VA TASVIRLARNI SHOVQINDAN TOZALASHDA MASHINAVIY O‘RGANISH USULLARI

##plugins.themes.bootstrap3.article.sidebar##

PDF (Русский)
Nashr: Jild 3 Nomeri 53 : Молодые ученые
Bo'lim: Articles
##semicolon##

##plugins.themes.bootstrap3.article.main##

Universitet: Farg’ona Davlat Texnika Universiteti
alisherbekdehqonov43@gmail.com
Universitet: Farg’ona Davlat Texnika Universiteti
aminjonovnurulloh25@gmail.com

Abstrak:

Ushbu maqolada signal va tasvirlarni shovqindan tozalashga qaratilgan mashinaviy o‘rganish usullari o‘rganildi. Konvolyutsion neyron tarmoqlar (CNN), avtrokodlovchi (Autoencoder), rekurrent neyron tarmoqlar (RNN, LSTM) va transformer arxitekturasi asosida qurilgan modellar tahlil qilindi. Har bir yondashuvning afzalliklari, cheklovlari va amaliy qo‘llanish sohalari ko‘rsatildi. Shuningdek, turli modellarning tasvir va audio signalni tozalashdagi samaradorligi bo‘yicha tajriba natijalari taqdim etildi.

##plugins.themes.bootstrap3.article.details##

##submission.howToCite##:

Dexqonov , A., & Aminjonov , N. (2025). SIGNAL VA TASVIRLARNI SHOVQINDAN TOZALASHDA MASHINAVIY O‘RGANISH USULLARI. Молодые ученые, 3(53), 99–101. Retrieved from https://in-academy.uz/index.php/yo/article/view/69008

##submission.citations##:

Goodfellow I., Bengio Y., Courville A. Deep Learning, MIT Press, 2016.

Ronneberger O. U-Net Architecture, MICCAI, 2015.

Zhang K. DnCNN: Beyond Gaussian Denoiser, IEEE CVPR, 2017.

Vaswani A. Attention Is All You Need, NIPS, 2017.

Pascual S. SEGAN: Speech Enhancement GAN, INTERSPEECH, 2017.

Luo Y. & Mesgarani N. Conv-TasNet, ICASSP, 2019.