MUMBAI, India, Feb. 27 -- Intellectual Property India has published a patent application (202641017322 A) filed by Easwari Engineering College; and Srm Institute Of Science And Technology, Chennai, Tamil Nadu, on Feb. 17, for 'deepclean: intelligent image de-noising using gan-based deep learning models for enhanced visual quality.'
Inventor(s) include Magima S; Manish Maadhav R S; Nithin S; and Mrs. M. Gowthami.
The application for the patent was published on Feb. 27, under issue no. 09/2026.
According to the abstract released by the Intellectual Property India: "This invention presents DeepClean, an intelligent deep learning-based image denoising framework designed to recover high-quality visual information from noise-corrupted images used in computer vision and medical imaging. The system utilizes a Generative Adversarial Network (GAN) with a convolutional encoder-decoder generator and a multi-scale discriminator to remove complex noise while retaining fine structural details. A hybrid loss function combining adversarial, perceptual, and L1 reconstruction losses ensures a balance between visual realism and pixel-level accuracy, reducing over-smoothing effects seen in traditional methods. DeepClean is trained on datasets containing synthetic Gaussian and Poisson noise along with real sensor-induced distortions, enabling strong generalization across varied imaging conditions. Experimental results show improved PSNR and SSIM compared to existing CNN- and Transformer-based denoising models. The framework supports real-time, high-resolution processing and is applicable to medical imaging, fracture detection preprocessing, low-light enhancement, satellite imagery, and image restoration."
Disclaimer: Curated by HT Syndication.
