A Study on the Performance of Adaptive Neural Networks for Haze Reduction with a Focus on Precision DOI Creative Commons
Ahmed Alshahir, Khaled Kaâniche, Ghulam Abbas

et al.

Mathematics, Journal Year: 2024, Volume and Issue: 12(16), P. 2526 - 2526

Published: Aug. 15, 2024

Visual clarity is significantly compromised, and the efficacy of numerous computer vision tasks impeded by widespread presence haze in images. Innovative approaches to accurately minimize while keeping image features are needed address this difficulty. The difficulties current methods need create better ones brought light investigation removal problem. main goal provide a region-specific reduction approach utilizing an Adaptive Neural Training Net (ANTN). suggested technique uses adaptive training procedures with external images, pixel-segregated haze-reduced Iteratively comparing spectral differences hazy non-hazy areas improves accuracy decreases errors. This study shows that recommended strategy upon existing ratio, region differentiation, precision methods. results demonstrate proposed method effective, 9.83% drop mistake rate 14.55% differentiating time. study’s findings highlight value adaptable neural networks for without losing quality. research concludes positive outlook on future methods, which should lead visual overall performance across wide range applications.

Language: Английский

Gradual task complexity scaling (GTCS-DRL): a deep reinforcement learning approach for training automated guided vehicle system DOI

Mohamed Rhazzaf,

Tawfik Masrour

Evolving Systems, Journal Year: 2025, Volume and Issue: 16(2)

Published: Feb. 20, 2025

Language: Английский

Citations

0
A Study on the Performance of Adaptive Neural Networks for Haze Reduction with a Focus on Precision DOI Creative Commons
Ahmed Alshahir, Khaled Kaâniche, Ghulam Abbas

et al.

Mathematics, Journal Year: 2024, Volume and Issue: 12(16), P. 2526 - 2526

Published: Aug. 15, 2024

Visual clarity is significantly compromised, and the efficacy of numerous computer vision tasks impeded by widespread presence haze in images. Innovative approaches to accurately minimize while keeping image features are needed address this difficulty. The difficulties current methods need create better ones brought light investigation removal problem. main goal provide a region-specific reduction approach utilizing an Adaptive Neural Training Net (ANTN). suggested technique uses adaptive training procedures with external images, pixel-segregated haze-reduced Iteratively comparing spectral differences hazy non-hazy areas improves accuracy decreases errors. This study shows that recommended strategy upon existing ratio, region differentiation, precision methods. results demonstrate proposed method effective, 9.83% drop mistake rate 14.55% differentiating time. study’s findings highlight value adaptable neural networks for without losing quality. research concludes positive outlook on future methods, which should lead visual overall performance across wide range applications.

Language: Английский

Citations

1