MUMBAI, India, Feb. 27 -- Intellectual Property India has published a patent application (202541133985 A) filed by Dr. Mahalingam Collegeo Of Engineering And Technnology, Pollachi, Tamil Nadu, on Dec 31 for 'texture-invariant malware detection using deep learning gan network.'
Inventor(s) include Dr. S. Ramakrishnan; and Mr. N. Praveen Sundra Kumar.
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: "Malware is malicious program which is harm to our computer system. In the recent era, different types ofmalware came into existence. Growth ofmalware is enormous with recent era new technologies. So finding new malware is a challenging task. Though there are various methods available to detect the malware, detection of texture invariant malware is a challenging. The present invention is a method for enharcing the performance of mal ware detection using a deep neural network. This method utilizes the Mallmg dataset, comprising 25 different families of rnalware images. Initially, the dataset undergoes preprocessing, where images are converted to grayscale and scaled. Features are extracted using a two-stage autoencoder process. The first stage autoencoder extracts features with a hidden layer size of I 0, and the second stage with a hidden layer size of 500. These features are then classified using a softmax layer. Additionally, statistical features such as entropy and energy are extracted using GLCM. The effectiveness of the proposed method is evaluated through performance metrics like accuracy, precision, recall, and F I score. Finally, this method demonstrates high specificity and efficiency in detecting various types of mal ware, offering a robust solution for mal ware detection in computer systems."
Disclaimer: Curated by HT Syndication.
