Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Dimitris Ampeliotis has a total of 39 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Synthetic Face Discrimination via Learned Image Compression

    by Sofia Iliopoulou, Panagiotis Tsinganos, Dimitris Ampeliotis, Athanassios Skodras

    The emergence of deep learning has sparked notable strides in the quality of synthetic media. Yet, as photorealism reaches new heights, the line between generated and authentic images blurs, raising concerns about the dissemination of counterfeit or manipulated content online. Consequently, there is a pressing need to develop automated tools capable of effectively distinguishing synthetic images, especially those portraying faces, which is one of the most commonly encountered issues. In this work, we propose a novel approach to synthetic face discrimination, leveraging deep learning-based image compression and predominantly utilizing the quality metrics of an image to determine its authenticity.


  • Could Human Gaze Augment Detectors of Synthetic Images?

    by Nikolaos Fotopoulos, Clara Riedmiller, Efe Bozkir, Panagiotis Tsinganos, Dimitris Ampeliotis, Gjergji Kasneci, Enkelejda Kasneci, Athanassios Skodras

    Recent advances in generative adversarial networks (GANs) allow for the synthesis of extremely photo-realistic face images, deceiving even the most experienced observers, let alone the unsuspecting internet user. Due to this, there has been a considerable effort by the image forensics community to design appropriate tools for the detection of these images. This paper first implements one such detection technique based on spatial and cross-band co-occurrence matrices and convolutional neural networks (CNNs), and then attempts to improve it by introducing additional information obtained from the human gaze. We show that in cases where human observers correctly decide whether an image is real or fake, eye movement information in combination with spatial and cross-band co-occurrence matrices derived from observation regions can be informative towards the task of detecting fake images. However, only a limited increase in the detection accuracy is achieved.

  • Connections Between Deep Equilibrium and Sparse Representation Models With Application to Hyperspectral Image Denoising

    by Alexandros Gkillas, Dimitris Ampeliotis, Kostas Berberidis

    In this study, the problem of computing a sparse representation of multi-dimensional visual data is considered. In general, such data e.g., hyperspectral images, color images or video data consists of signals that exhibit strong local dependencies. A new computationally efficient sparse coding optimization problem is derived by employing regularization terms that are adapted to the properties of the signals of interest. Exploiting the merits of the learnable regularization techniques, a neural network is employed to act as structure prior and reveal the underlying signal dependencies. To solve the optimization problem Deep unrolling and Deep equilibrium based algorithms are developed, forming highly interpretable and concise deep-learning-based architectures, that process the input dataset in a block-by-block fashion. Extensive simulation results, in the context of hyperspectral image denoising, are provided, which demonstrate that the proposed algorithms outperform significantly other sparse coding approaches and exhibit superior performance against recent state-of-the-art deep-learning-based denoising models. In a wider perspective, our work provides a unique bridge between a classic approach, that is the sparse representation theory, and modern representation tools that are based on deep learning modeling.


  • Open Access

    (INVITED)Review of Optical Fiber Technologies for Optogenetics

    by Anastasios Tsakas, Christos Tselios, Dimitris Ampeliotis, Christina (Tanya) Politi, Dimitris Alexandropoulos


  • Open Access

    Proximal Algorithms for Discrete-Level Phase-Shifting Mask Design with Application to Optogenetics

    by Dimitris Ampeliotis, Aggeliki Anastasiou, Christina (Tanya) Politi, Dimitris Alexandropoulos

    This work studies the problem of designing computer-generated holograms using phase-shifting masks limited to represent only a small number of discrete phase levels. This problem has various applications, notably in the emerging field of optogenetics and lithography. A novel regularized cost function is proposed for the problem at hand that penalizes the unfeasible phase levels. Since the proposed cost function is non-smooth, we derive proper proximal gradient algorithms for its minimization. Simulation results, considering an optogenetics application, demonstrate that the proposed proximal gradient algorithm yields better performance as compared to other algorithms proposed in the literature.


  • Open Access

    Potential Games for Distributed Constrained Consensus

    by Dimitris Ampeliotis, Kostas Berberidis

    The problem of computing a common point that lies in the intersection of a finite number of closed convex sets, each known to one agent in a network, is studied. This issue, known as the distributed convex feasibility problem or the distributed constrained consensus problem, constitutes an important research goal mainly due to the large number of possible applications. In this work, this issue is treated from a game theoretic viewpoint. In particular, we formulate the problem as a non-cooperative game for which a potential function exists and prove that all Nash equilibria of this game correspond to consensus states. Based upon this analysis, a best-response based distributed algorithm that solves the constrained consensus problem is developed. Furthermore, one more approach to solve the convex feasibility problem is studied based upon a projected gradient type algorithm that seeks the maximum of the considered potential function. A condition for the convergence of this scheme is derived and an exact distributed algorithm is given. Finally, simulation results for a source localization problem are given, that validate the theoretical results and demonstrate the applicability and performance of the derived algorithms.

  • Distributed Set-Theoretic Parameter Estimation in Networks with Ambiguous Measurements

    by Dimitris Ampeliotis, Christos Mavrokefalidis, Kostas Berberidis, Sergios Theodoridis

    Distributed estimation of a parameter vector in a network of sensor nodes with ambiguous measurements is considered. These ambiguities may be due to interference, poor calibration, high noise levels or any other cause. Cooperation among the nodes is required to resolve them. In such a setting, non-convex constraint sets may be required at the nodes, in order to accurately model the local ambiguities. We propose to handle the resulting non-convexity by expressing the involved non-convex sets as unions of convex sets, such that, for each node, only one such convex set is actually relevant to the estimation task. A procedure is employed in which the nodes properly select one of their convex sets with the aim to reach network-wide agreement. In this formulation, agreement can be reached when the nodes select sets with non-empty intersection. To this end, a learning algorithm is proposed and then a sketch of a proof for its convergence under some mild conditions is carried out. Finally, proper numerical results that support the theoretical findings are shown.


  • Adapt-Align-Combine for diffusion-based distributed dictionary learning

    by D. Ampeliotis, C. Mavrokefalidis, K. Berberidis, S. Theodoridis

    Diffusion-based distributed dictionary learning methods are studied in this work. We consider the classical mixed l2-l1 cost function, that employs an l2 representation error term and an l1 sparsity promoting regularizer. First, we observe that this cost function suffers from an inherent permutation ambiguity. This ambiguity may deteriorate significantly the performance of diffusion-based schemes, since the involved combination step may combine different atoms even when the same atoms exist at all dictionaries. Thus, we propose to align the dictionaries prior to the combination step. Furthermore, we define a new problem, that we call the node-specific distributed dictionary learning problem. The proposed Adapt-Align-Combine algorithm enjoys increased convergence rate as compared with a scheme that does not align the dictionaries prior to the combination. Simulation results support our findings.