Chhattisgarh Swami Vivekanand Technical University

Composite materials based on polymers are being identified as helping to promote sustainable infrastructure by providing lightweight, durable, and corrosion-resistant solutions. Their applications span construction, transportation, energy, and marine sectors, enabling longer service life, reduced maintenance, and enhanced performance of modern structural systems. The curing process plays an essential role in determining the performance and reliability of polymer composites, influencing their mechanical, thermal, and chemical properties. This review aims to explore the advances, challenges, and opportunities in curing methods for polymer composites, focusing on both traditional and emerging techniques. Key motivations include the need for faster curing times, improved material properties, and enhanced sustainability in the processing of composite materials. The review covers all forms of curing, ranging from heat, ultraviolet, and microwave curing to more recent technologies like additive manufacturing. It further addresses existing challenges, including process optimization in curing, enhancement of energy efficiency, and ensuring compatibility across materials. Various research highlights that significant progress has already been made, still there is a need to overcome current challenges and design the full potential of polymer composites for the next-generation market. The review also explores sustainable curing approaches that reduce environmental impact by minimizing energy consumption and incorporating eco-friendly additives, thus advancing the development of green polymer composites for building and construction applications.

Ensuring public safety in densely populated urban environments remains a critical challenge, necessitating the deployment of intelligent and automated video surveillance systems. Traditional surveillance approaches rely heavily on manual monitoring, which is inefficient and susceptible to human fatigue, delayed response, and observational errors. To overcome these limitations, this work presents a real-time object detection-based surveillance framework. The proposed system focuses on detecting guns, knives, and region-specific blunt objects commonly involved in violent activities in Indian surveillance scenarios. A key contribution of this work is the use of a custom-created dataset collected using a mobile camera, consisting of 336 labeled images of blunt objects such as iron rods, wooden sticks, and plastic rods. This dataset is combined with a publicly available dataset of 7,623 images of guns and knives, forming a consolidated dataset of 7,959 images across three classes: gun, knife, and blunt object. The combined dataset is used to train a YOLOv8-based object detection model for real-time performance. Experimental evaluation shows that increasing the training duration significantly improves recall and average precision for the blunt object class without signs of overfitting. Overall, the proposed framework achieves an effective balance between accuracy and efficiency, making it suitable for deployment in real-world surveillance environments such as campuses, public spaces, and transportation areas.

Sarcasm detection has become an essential task in natural language processing, especially in social media, where sarcasm is frequently used to convey nonliteral meanings that can distort sentiment analysis outcomes. This study proposes a multimodal approach that integrates text and audio features to enhance sarcasm detection accuracy. Traditional sarcasm detection methods often rely solely on text-based features, which may fail to capture the vocal cues associated with sarcastic intent. Our proposed model combines text features such as TF-IDF, word embeddings, and sentiment polarity shifts with audio features, including pitch, energy intensity, and speech rate. By using a Support Vector Machine (SVM) classifier and optimizing the hyperparameters, we achieved an accuracy of 87.8

In many computer vision applications such as intelligent transportation systems, autonomous navigation, digital content retrieval, and multilingual translation, detection and recognition of scenes play a vital role. These systems are fairly effective at document-oriented applications but less effective in real-world applications where an increased number of languages, orientation variation, noise, and distortion are present. To overcome these shortcomings, we are introducing a Hybrid Structured Text Extraction Network (HSTENet) for robust and accurate scene-text recognition that combines deep learning and machine learning. HSTENet features advanced modules such as edge detection with image preprocessing, support vector machine classification for detecting the textual part, and a line opening algorithm for fast segmentation of the image into regions. The recognition part relies on deep architecture, convolutional neural networks, attention-based connectionist temporal classification decoding, and bidirectional long short-term memory networks to recognize the high-precision target text. HSTENet was evaluated on three benchmark datasets—the International Conference on Document Analysis and Recognition, Common Objects in Context for Text, and a self-designed high-noise multilingual dataset—and obtained an F1-score of 91.9% and recognition accuracy of 93.8%, significantly outperforming other state-of-the-art models, including DeepText, convolutional recurrent neural networks, and connectionist text proposal networks. The framework is also polarization-invariant, real-time compatible (inference time 0.69 s), and deployable on edge devices. The results showed that HSTENet is highly efficient and automated for textual scene recognition, which could be applied to systems embedded in mobile platforms, self-contained platforms, and intelligent human–computer interaction.

With the rising cases of mental conditions like stress, anxiety, and depression in modern societies, there has been a great need for automated and contactless systems to deal with these issues. Emotional and psychological aspects in speech are full of important cues that can be tapped to detect an emotional imbalance or stress in people. Traditional systems based on speech emotion recognition use acoustic features that lack universality in terms of speakers and environments.This paper proposes a hybrid deep learning architecture for joint emotion recognition and stress estimation from raw audio signals. The suggested approach combines a Wav2Vec2 self-supervised speech representation model pretrained on raw audio data with a BiLSTM architecture designed to extract long-term dependencies. For multitask learning, two distinct output layers are used for emotion recognition and stress estimation.The training and testing of this proposed model is performed on a combined dataset created from RAVDESS, TESS, and SAVEE speech databases after performing label normalization on them. The experimental results of this proposed approach prove that it is effective since it achieves an accuracy of 85.9% in categorizing emotions and a stress value prediction RMSE of 0.1268. It demonstrates potential for deployment in mental health monitoring applications.