Ramrao Adik Institute of Technology

Corrosion of metals in diverse media degrades their functional and operational properties. After corrosion, metals are not as useful as they are in pristine conditions. Mild steel (MS) is a highly exploited metal in engineering works due to its supreme properties. Several methods have been adopted by the scientists to prevent the corrosion of MS; however, the prevention by natural products as corrosion inhibitors is more appreciated and needed due to environmental concerns. However, it is very important to know the inhibition mechanism of inhibitors before using them in prevention applications, especially in acids. This review gives a brief review of corrosion inhibitors and focuses on their working mechanism for MS in strong acids. The investigation has been done by analysing the articles published on the use of natural materials for MS’s protection in acid media. The results of weight loss measurements (WLM), electrochemical measurements (ECM), surface analysis, and computational techniques in those articles were analysed and the inhibition mechanism has been discussed. This review article provides a clear and broad view of corrosion inhibition mechanisms and also gives suggestions to enhance the performance, which could be beneficial for the scientists and researchers working in this area.

Audio source separation (ASS) is a technique well-known for extracting the individual signal of underlying sound sources from the signal mixture, which is a hectic challenge due to the presence of multi-channel signals. Though there are numerous features associated with the existing separation models, time-domain signals and phase information are not considered in the conventional techniques. On the other hand, separation models working with DNN are not so efficient with sampling and the separation accuracy of the signal. To overcome these drawbacks, the optimized U-Net model with the Hybrid Wolf Optimization (HWO-UNet) is proposed in this research, which aids in separating the audio source with better accuracy. Utilizing the Hybrid optimized U-Net model with skip connection integrates the spectrogram features and the statistical features effectively for eliminating the loss of signal features and assists in the effective separation of the audio signal as well as the noise signal. The proposed method adopts the U-NET-based architecture with optimized intermediate spectrogram transformation blocks utilizing the adaptive tuning behavior of Deep CNN. Further, the U-Net acts as an encoder–decoder architecture that effectively works as an oversampling technique that mitigates the class imbalance problem and the anti-aliasing technique enhances the efficiency. The performance of the HWO-UNet ASS method in terms of Root Mean Square Error (RMSE), and Mean Square Error (MSE) is 1.242, and 0.912, respectively for the MUSDB18 whereas the maximal Signal-to-Noise Ratio (SNR) is reported as 16.51 dB for UrbanSound8k dataset.

With rapid advances in Deep Learning (DL) algorithms and social media content, deepfakes have emerged as a potent tool for manipulating multimedia content to commit defamation, falsify information, and pose other security threats. Recent deepfake detection is challenging due to the intricate structure of deep learning models, high overlap between real and fake data, lower reliability, limited generalization, and poor interpretability and explainability of deepfake detection models. This paper presents a robust and reliable DeepFake Detection Network (DeepFakeDetNet) for videos using a Deep Convolutional Neural Network (DCNN) and Long Short-Term Memory (LSTM) networks to improve the generalization capability, interpretability, accuracy and minimize the computation intricacy of the model. The system considers the Image Texture and Shape Features (ITSF) that encompass the Gray level Co-occurrence Matrix (GLCM) feature to depict the spatial relationship in texture, a novel Extended Local Ternary Pattern (XLTP) to provide local texture patterns, and Histogram of Oriented Gradients (HOG) to depict the shape attributes of the deepfake images. Multiple Acoustic Features (MAFs) are used to depict the spectral, temporal, and phonetic attributes of the audio. Furthermore, an Improved Starfish Optimization Algorithm (ISOA) is employed for feature selection, focusing on prominent features to reduce the system’s computational complexity. The effectiveness of the proposed deepfake detection system is evaluated on the FakeAVCeleb dataset. The proposed ITSF + MAF+SOA offers better generalization capability, superior spectral-temporal depiction of multimodal deepfake modalities, and minimization in computational intricacy of the deepfake detection system. The ITSF + MAF+ISO achieves an improved accuracy of 97.80

Understanding high spatial resolution (HSR) remote sensing (RS) imagery requires exploring geo-objects and their geographic relationships. This area has grabbed the interest of the RS community since it delivers more specific information than typical tasks like classification and object recognition. Despite recent advancements in RS image description generation, it remains challenging to characterize the RS image in terms of geographic relationships between the objects included in it. This work proposes a method for semantic understanding of high spatial resolution RS images by identifying topological, directional, and proximity geospatial relationships between geo-objects and representing these relationships in the form of sentences. The proposed methodology began with the detection of RS image objects in the form of oriented bounding boxes (OBB). The dimensionality extended 9-intersection model (DE9IM) was then used to identify topological relations. Directional relationships are then identified using the centroid of both objects, and proximity relations are identified based on the distance between two objects. Thus, between each pair of objects, all three types of relations are constructed, and relation triplets have been created for each relation. To decrease the redundancy in the relationships, a hierarchy of relations is proposed, and an extended region connection calculus (ERCC) for nesting of relations is constructed. ERCC is generated by linking directional and proximity relations with topological relations, and three levels of relations are designed. The methodology is validated using a 34-class RS image dataset constructed for object detection in the form of oriented bounding boxes.

Underwater images play a significant role in numerous marine applications, including underwater archaeology, ocean exploration, and marine ecological research. However, the underwater images are suffered with the quality degradation due to scattering effects, light absorption, and low illumination. Thus, effective Underwater Image Enhancement (UIE) approaches are vital to increase the visibility and visual information of the images. The traditional enhancement approaches are limited in capturing the global contextual information and generate over-enhanced images, and also exhibit higher computational cost and complexity. To address these limitations, this research proposed a U-shape Transformer Diffusion Denoising (U-shape TD2) model to enhance the underwater images. The U-shape TD2 model integrates the strength of the transformer model and the diffusion-based denoising process within a U-shaped architecture, which empowers the model to capture the global contextual information, thereby eliminating the noise and enhancing the visual content. Overall, the U-shape TD2 model increases the quality and preserves the significant visual information. The experimental outcomes illustrate that the U-shape TD2 model attained the Structural Similarity Index (SSIM), Peak Signal-to-Noise Ratio (PSNR), and Feature Similarity Index (FSIM) values of 0.959, 57.96dB, and 0.865 using the LSUI dataset, respectively.