10.46243/jst.2023.v8.i12.pp07-16 registered

This paper presents an approach for changing the facial coordinates of a person in the video as per the given input image. To get the desired outcome we use many machine learning and deep learning algorithms like Generative adversarial network and auto encoders to manipulate the video facial expressions. It stands out as a pivotal feature, where we intend to develop a robust algorithm capable of seamlessly replacing one individual’s face with another while preserving the original image’s lighting conditions, facial expressions, and overall realism. This feature has vast potential for fun and entertainment, as well as applications in the film and advertising industries. 1,2,3B. Tech Student, Department of CSE (Cyber Security), Malla Reddy College of Engineering and Technology, Hyderabad,India. 4Professor and HoD,Department of CSE Emerging Technologies,Malla Reddy College of Engineering and Technology,Hyderabad,India. DOI:https://doi.org/10.46243/jst.2023.v8.i12.pp7-16 6 Sankeerth Reddy Jakkidi, Ajoji Aravind Kumar, Boompagul Pranay Kumar, Dr M V Kamal: FACE CHANGER USING DEEP FAKE IN PYTHON This is mainly used to do some funny things and fancy computer tricks to do cool stuff with people’sfaces in pictures and videos. Additionally, our project aims to delve into the realm of emotion recognition. By harnessing state-of-the-art deep learning techniques, we plan to build a facial emotion recognition system that can accurately detect and analyze emotions displayed on human faces. This could have profound implicationsin psychology, enabling researchers to gain insights into emotional responses, and also benefit market research by gauging consumer reactions to products or advertisements. In an era defined by rapid advancements in artificial intelligence and computer vision, the convergence of deep learning techniques and real-time video manipulation has given rise to an innovative technology known as DeepFace Live. This groundbreaking concept represents a new frontier in visual media manipulation, enabling the seamless and dynamic alteration of facial expressions, features, and even entire identities. Deepfake methods normally require a large amount of image and video data to train models to create photo-realistic images and videos. In today’s digital era, the realm of image and video manipulation has witnessed a remarkable evolution, thanks to the advent of deep learning techniques. Among the most intriguing and, at times, controversial innovations in this field is the concept of deep fakes. These sophisticated neural networks have the power to seamlessly alter the faces of individuals in images and videos, ushering in a new era of creative expression, entertainment, and, simultaneously, raising critical ethical considerations. The “Face Changer using Deep Fake in Python” project is a fascinating exploration of this groundbreaking technology, offering a practical and responsible tool for facial transformation. This project harnesses the potential of deep learning, typically employing advanced models like Generative Adversarial Networks (GANs) or Variational Autoencoders (VAEs), to achieve realistic and convincing facial alterations. With a focus on high-quality data preparation, the model is trained on diverse datasets encompassing various facial expressions, lighting conditions, and angles, ensuring robust performance. What sets this project apart is its commitment to user-friendliness. Through a Python-based interface, users can effortlessly upload images or videos and select their desired facial transformations, such as altering identity, expression, or age. Real-time processing capabilities further enhance the user experience, making it possible to apply facial changes within live video streams. Nevertheless, the ethical implications of deep fake technology are not taken lightly. The project places a strong emphasis on ethical considerations, including the potential for misuse. It includes safeguards and disclaimers to promote responsible usage and user education. The pursuit of quality and performance is paramount, with ongoing efforts to fine-tune the model and optimize processing speed. Security and privacy are also fundamental aspects of the project, with measures in place to prevent unauthorized manipulation and protect individuals from misuse. LITERATURE SURVEY Based on the observation that temporal coherence is not enforced effectively in the synthesis process of deep- fakes, Sabir et al. [103] leveraged the use of spatio-temporal features of video streams to detect deepfakes. Video manipulation is carried out on a frame-by-frame basis so that low level artifacts produced by face manipulations are believed to further manifest themselves as temporal artifacts with inconsistencies across frames.A recurrent convolutional model (RCN) was proposed based on the integration of the convolutional network DenseNet and the gated