Logo des Repositoriums

Auflistung nach Schlagwort "Generative Adversarial Networks"

1 - 4 von 4
  • Textdokument
    Achieving Facial De-Identification by Taking Advantage of the Latent Space of Generative Adversarial Networks
    (INFORMATIK 2021, 2021) Frick, Raphael Antonius; Steinebach, Martin
    The General Data Protection Regulation (EU)2016/679 passed by the European Union prohibits any data collection and processing that was conducted without the consent of the individuals involved. Especially images showing faces are often subject to these regulations and therefore, either need to be removed or anonymized. Early approaches however were often troubled by strong visual artifacts. In this work, we propose a novel anonymization pipeline that generates a proxy face for a group of individuals by taking advantage of the semantics of the latent space of generative adversarial networks. Experiments have shown that by following a 𝑘-same approach and utilizing different clustering techniques, privacy for the individuals involved can be greatly enhanced, while preserving important facial characteristics.
  • Textdokument
    Astronomical Image Colorization and Up-scaling with Conditional Generative Adversarial Networks
    (INFORMATIK 2022, 2022) Kalvankar,Shreyas; Pandit,Hrushikesh; Parwate,Pranav; Patil,Atharva; Kamalapur,Snehal
    This research aims to provide an automated approach for the problem of Image colorization and Single Image Super Resolution by focusing on a very specific domain: astronomical images, using Generative Adversarial Networks. We explore the usage of various models in RBG and L*a*b color spaces. We use transfer learning owing to a small data set, using pre-trained ResNet-18 as a backbone encoder and fine-tune it further. The model produces visually appealing images that are high resolution and colorized. We present our results by evaluating the GANs quantitatively using distance metrics such as L1 distance and L2 distance in each of the color spaces across all channels to provide a comparative analysis. We use Fréchet inception distance (FID) to compare the distribution of the generated images and real image to assess the model's performance.
  • Konferenzbeitrag
    From attributes to faces: a conditional generative network for face genera-tion
    (BIOSIG 2018 - Proceedings of the 17th International Conference of the Biometrics Special Interest Group, 2018) Wang, Yaohui; Dantcheva, Antitza; Bremond, Francois
    Recent advances in computer vision have aimed at extracting and classifying auxiliary biometric information such as age, gender, as well as health attributes, referred to as soft biometrics or attributes. We here seek to explore the inverse problem, namely face generation based on attribute labels, which is of interest due to related applications in law enforcement and entertainment. Particularly, we propose a method based on deep conditional generative adversarial network (DCGAN), which introduces additional data (e.g., labels) towards determining specific representations of generated images. We present experimental results of the method, trained on the dataset CelebA, and validate these based on two GAN-quality-metrics, as well as based on three face detectors and one commercial off the shelf (COTS) attribute classifier. While these are early results, our findings indicate the method’s ability to generate realistic faces from attribute labels.
  • Konferenzbeitrag
    A Generalizable Deepfake Detector based on Neural Conditional Distribution Modelling
    (BIOSIG 2020 - Proceedings of the 19th International Conference of the Biometrics Special Interest Group, 2020) Khodabakhsh, Ali; Busch, Christoph
    Photo- and video-realistic generation techniques have become a reality following the advent of deep neural networks. Consequently, there are immense concerns regarding the difficulty in differentiating what content is real from what is synthetic. An example of video-realistic generation techniques is the infamous Deepfakes, which exploit the main modality by which humans identify each other. Deepfakes are a category of synthetic face generation methods and are commonly based on generative adversarial networks. In this article, we propose a novel two-step synthetic face image detection method in which general-purpose features are extracted in a first step, trivializing the task of detecting synthetic images. The anomaly detector predicts the conditional probabilities for observing every individual pixel in the image and is trained on pristine data only. The extracted anomaly features demonstrate true generalization capacity across widely different unknown synthesis methods while showing a minimal loss in performance with regard to the detection of known synthetic samples.