The family of compact composite descriptors includes descriptors for 3 types of images. A group of 3 descriptors combines color and texture information in order to describe natural color images. On this page, you can read more details about CCDs and download the related libraries.
From this page, you can download an open-source implementation (in C#) of the Scalable Color Descriptor, Color Layout Descriptor, Dominant Colors Descriptor, as well as the Edge Histogram Descriptor.
We employ the SURF detector to define salient image patches of blob-like textures and use the MPEG-7 Scalable Color (SC), Color Layout (CL), Edge Histogram (EH) and the Color and Edge Directivity Descriptor (CEDD) descriptors to produce the final local features’ vectors named SIMPLE-SC, SIMPLE-CL, SIMPLE-EH, and SIMPLE-CEDD or “LoCATE” respectively.
img(Rummager) brings into effect a number of new as well as state of the art descriptors. The application can execute an image search based on a query image, either from XML-based index ﬁles or directly from a folder containing image ﬁles, extracting the comparison features in real-time.
In this web-site, a new set of feature descriptors is presented in a retrieval system. These descriptors have been designed with particular attention to their size and storage requirements, keeping them as small as possible without compromising their discriminating ability.
The LIRE (Lucene Image REtrieval) library provides a simple way to retrieve images and photos based on their color and texture characteristics. LIRE creates a Lucene index of image features for content-based image retrieval (CBIR). Three of the available image features are taken from the MPEG-7 Standard: ScalableColor, ColorLayout, and EdgeHistogram a fourth one, the Auto Color Correlogram has been implemented based on recent research results.
Enorasis robot implements a computer-vision-based line-tracking and following control technique for high-speed autonomous navigation.
Olive oil is one of the major agricultural products of the countries of the Mediterranean Basin. Currently, more than 70% of the world’s production is produced in Spain, Italy, and Greece. One of the main threats for olive production is the pest infestation which results in the distraction of the fruit of the olive tree.
To tackle the problem of successful implementation of bait sprays we have developed a holistic approach using embedded systems, GIS and web-based applications
Research on Blockchain
Our research on Blockchain focuses on studying the innovative concepts leveraged by this newly emerged technology and by its new application-driven directions. Our main research activity focuses on designing and developing special-purpose decentralized applications (dApps).
Our work includes the Parkchain project, the RandomBlocks generator as wel as Sensitive Information Excange Mechanisms.
Edge Intelligence group
The Edge Intelligence group aims to research performance optimizations for improving network edge service support for IoT, IoV and tactile internet mobile devices. Depending on the research objective this may involve advances in a combination of several research areas at the Intelligent Edge including but not limited to mobile computing, wireless access networks, network performance, resource management, task offloading, load balancing, scheduling techniques, vehicle-to-infrastructure (V2I), Cyber-Physical Systems (CPS), network edge security and federated learning.
Research on Intelligent Network on Chip (NoC)
Our research on Intelligent NoCs aims to explore performance optimizations for improving networks on chip. Depending on the research objective this may involve advances in a combination of several research areas at the Networks on Chip (NoC) including but not limited to an exploration and development of high-throughput integrated interconnect network on chip through topologies, routing algorithms, flow control protocols using intra-silicon photonic links, which are a novel technology in the field of hybrid (photonic-electric) integrated interconnect networks on chip. As well as, a combination and use of advanced analytics technologies, including machine learning, deep reinforcement learning and predictive modelling, to support the identification of hotspots, fault tolerance and provide automated routing decisions etc., promises degradation in latencies and improvement of performance on NoCs.
Investigating the Vision Transformer Model for Image Retrieval Tasks
This work presents a new method that utilizes the recently proposed Vision Transformer network to shape a global descriptor. In image retrieval tasks, the use of Handcrafted global and local descriptors has been very successfully replaced, over the last years, by the Convolutional Neural Networks (CNN)-based methods. However, the experimental evaluation conducted in this paper demonstrates that a neural network that contains no convolutional layer, such as Vision Transformer, can shape a global descriptor and achieve competitive results.
A Gradient Free Neural Network Framework Based on Universal Approximation Theorem
This paper presents a numerical scheme for the computation of Artificial Neural Networks’ weights, without a laborious iterative procedure. The proposed algorithm adheres to the underlying theory, is highly fast, and results in remarkably low errors when applied for regression and classification of complex data-sets, such as the Griewank function of multiple variables x∈R^100 with random noise addition, and MNIST database for handwritten digits recognition, with 7×10^4 images.
The Lagrangian remainder of Taylor’s series, distinguishes O(f(x)) time complexities to polynomials or not
The purpose of this letter is to investigate the time complexity consequences of the truncated Taylor series, known as Taylor Polynomials. In particular, it is demonstrated that the examination of P=NP equality, is associated with the determination of whether the n^th derivative of a particular solution is bounded or not. Accordingly, in some cases, this is not true, and hence in general.