In present era there are various methods and different features which can be used to perform face recognition and emotion recognition like face, text, speech, and so on. The facial feature among these is the best. Emotion detection is also widely used in many fields like understanding human behaviour, detecting mental disorders, finding the emotions of people in a crowd, etc. In the proposed framework, face-emotion detection can be defined in three stages. The first stage is detecting the human face from the camera feed then in the second stage, the input is analysed based on the features with help of the convolutional neural networks (CNN) model. The last stage is to classify human emotion into 7 basic categories: fear, anger, disgust, sad, surprise, neutral and happy. Thus, proposed framework has mainly three objectives like face detection, face recognition and emotion classification. Face recognition has many significant applications in various fields like identity verification, security, biometrics, smart card, surveillance system, etc. Facial expression identification eminently comes from a few parts of the face, like eyes and mouth. The framework proposes to capture the facial features in real-time and process them to identify if there are any similarities between the real-time input and any of the trained faces. The outcome of the proposed framework is to display the name of the recognized individual and classify the face emotion with probabilities of the emotion. It also displays the emotion of an unknown person with the name given as ‘unidentified’, along with the probabilities of the emotions.
Addressing the intricacies of facial aging in forensic facial recognition, traditional sketch portraits often fall short in precision. This study introduces a pioneering system that seamlessly integrates a de-aging module and a sketch generator module to overcome the limitations inherent in existing methodologies. The de-aging module utilizes a deepfake-based neural network to rejuvenate facial features, while the sketch generator module leverages a pix2pix-based Generative Adversarial Network (GAN) for the generation of lifelike sketches. Comprehensive evaluations on the CUHK and AR datasets underscore the system’s superior efficiency. Significantly, comprehensive testing reveals marked enhancements in realism during the training process, demonstrated by notable reductions in Frechet Inception Distance (FID) scores (41.7 for CUHK, 60.2 for AR), augmented Structural Similarity Index (SSIM) values (0.789 for CUHK, 0.692 for AR), and improved Peak Signal-to-Noise Ratio (PSNR) metrics (20.26 for CUHK, 19.42 for AR). These findings underscore substantial advancements in the accuracy and reliability of facial recognition applications. Importantly, the system, proficient in handling diverse facial characteristics across gender, race, and culture, produces both composite and hand-drawn sketches, surpassing the capabilities of current state-of-the-art methods. This research emphasizes the transformative potential arising from the integration of de-aging networks with sketch generation, particularly for age-invariant forensic applications, and highlights the ongoing necessity for innovative developments in de-aging technology with broader societal and technological implications.
In today's era, unreliable news results in several consequences impacting the society changing the people's perspective. There are several approaches and techniques for detecting the fake news, also new methods are being implemented in order to obtain more accurate model. This paper focuses on developing a machine learning model using Natural Language Processing techniques like TF-IDF feature extraction and passive aggressive classifier to detect the fake news. The experimental outcomes shows that the proposed model outcomes existing model in detection of fake news.
The current era is driven by online social networks which have evolved as a popular virtual communication platform among people. Social media users often share pictures, videos and express their views on various matters through social media posts. Though online social networks have succeeded in helping people to interact and build connections worldwide virtually, it also exposes the user to different kinds of cyber frauds. Hence identifying a genuine user profile on social media has gained significant importance in the lieu of detecting social media users from cyber criminals. With this regard, this paper is focussed at developing a machine learning model that identifies and classifies user profiles as genuine or not genuine categories. The experiment outcome shows that the proposed model achieved an average accuracy of 95% in the classification task considering all three datasets.
For banks and other financial institutions, loans are the primary source of income. As a result, a variety of loan types are offered depending on certain limitations. Banks face difficulties in determining whether or not to approve the applicant's loan. To evaluate whether a loan application may be approved, loan data must be analyzed. Giving a loan without first analyzing the applicant's data is risky from the bank's perspective. In this study, loan data is evaluated utilizing a combination of feature selection and appropriate machine learning algorithms, including clustering, association, and classification methods. We can determine whether a loan can be approved or not by employing a number of algorithms. In order to choose the best strategy, these algorithms are also contrasted.
A city longing to get tagged as a smart city should always aim at the health and public safety of the citizens. Every initiative carried out by the city needs to confine itself with a boundary of safety. The boundary of safety can be defined with the aid of artificial intelligence. In a broad view, a clean and hygienic environment can be achieved by tagging the requirements with deep learning in every single application criterion. Cleanliness of city streets is a significant influence on the urban environment and healthy living. Standard street cleaning involves street sweepers going to clean several places by checking whether to clean or not. Manually detecting and assessing the cleanliness of street takes to a higher cost for cities. To overcome this vulnerability, noTrash.ai -an approach towards the hygienic city uses deep learning to stop further littering by monitoring the litters thrown from the vehicle in real time to classify the trash thrown from moving vehicles.