Figure 2 - uploaded by Thomas Di Martino
Content may be subject to copyright.
Source publication
State of the art models for Similarity Learning are all based on Deep Learning architecture using Siamese Network [Gregory et al., 2015]. They define a feature extraction pipeline that creates a latent representation of input data. This embedding vector is semantically highly descriptive and can be used for the computation of distances between data...
Similar publications
Automatic readability assessment (ARA) is the task of evaluating the level of ease or difficulty of text documents for a target audience. For researchers, one of the many open problems in the field is to make such models trained for the task show efficacy even for low-resource languages. In this study, we propose an alternative way of utilizing the...
Text classification has a growing interest among NLP researchers due to its tremendous availability on online platforms and emergence on various Web 2.0 applications. Recently, text classification in resource-constrained languages has been bringing much attention due to the sharp increase of digital resources. This paper presents a CNN based text c...
![Evaluate on Single-Human Optical Flow dataset (SHOF) [48]. Our method...](profile/Ruofei-Du-4/publication/350483762/figure/tbl1/AS:1006899319291904@1617074913261/Evaluate-on-Single-Human-Optical-Flow-dataset-SHOF-48-Our-method-achieve-the-best_Q320.jpg)
In this paper, we address the problem of building dense correspondences between human images under arbitrary camera viewpoints and body poses. Prior art either assumes small motion between frames or relies on local descriptors, which cannot handle large motion or visually ambiguous body parts, e.g., left vs. right hand. In contrast, we propose a de...
![Fig. 1. Example imagery of the Oxford RobotCar Dataset [21], showing...](profile/Ciaran-Eising/publication/349703937/figure/fig1/AS:996747971162113@1614654643213/Example-imagery-of-the-Oxford-RobotCar-Dataset-21-showing-different-weather-and_Q320.jpg)
Keypoint detection and description is a commonly used building block in computer vision systems particularly for robotics and autonomous driving. Recently CNN based approaches have surpassed classical methods in a number of perception tasks. However, the majority of techniques to date have focused on standard cameras with little consideration given...
Compared to feature point detection and description, detecting and matching line segments offer additional challenges. Yet, line features represent a promising complement to points for multi-view tasks. Lines are indeed well-defined by the image gradient, frequently appear even in poorly textured areas and offer robust structural cues. We thus here...
Citations
... Essentially, it measures the degree of similarity between images and outputs a score between 0 and 1, with higher values signifying greater contextual similarity. A score of 1 represents identical images (Ma et al., 2021) (Martino, 2020). Figure 3 illustrates the typical deep similarity learning process. ...
... Finally, based on the chosen metric, a similarity score is generated, reflecting the degree of similarity between the vectors. Higher scores signify a closer match between the original images, while lower scores indicate disagreement (Ma et al., 2021) (Martino, 2020). In the development of the Khatti application, image similarity processing was addressed by leveraging the availability of free and open-source software resources. ...
... Data processing pipeline for similarity learning adapted from(Martino, 2020). ...
Arabic calligraphy serves as a powerful symbol of cultural identity and heritage, evolving and adapting to modern contexts, contributing to a dynamic and vibrant cultural landscape. The issue of handwritten recognition and training training in Arabic script nature has attracted many researchers from both academic and industrial fields. Khatti, an innovative mobile application, was proposed to facilitate the training and learning of Arabic calligraphy. It can help the trainee to improve their calligraphy using technology-driven approach that utilizes interactive features to clarify the details of script, fostering an intuitive learning experience, guiding beginners through the fundamentals and supporting seasoned calligraphers in honing their artistry. This paper investigates the conceptual framework and technical aspects of Khatti, highlighting its potential to revolutionize calligraphy education with the integration of machine learning algorithms. Additionally, application's potential for future development and its possible applications in artistic research and cultural preservation was discussed.
