1,720,960 research outputs found
Latent Space Dynamics and Security Implications in Machine Learning for Industrial Systems
Full text linkL'abstract è presente nell'allegato / the abstract is in the attachmen
Defect Detection In Vehicle Painting: Case Study
Full text linkDefect detection is a cross-sectoral problem that is being intensively addressed in manufacturing, primarily
with the help of computer vision and image processing-based systems. From fabric to surface to mechanical
parts, defect detection approaches have assisted human operators and reduced human eye strain. However, many case-specific challenges arise in vehicle painting. Although few authors have addressed them, research is still active due to the high-quality demand and competition in manufacturing. In this study, we present a case study on paint defect detection in IVECO vehicle production, listing the problem description, challenges, literature review, and proposed solution
Exploring Latent Space Using a Non-linear Dimensionality Reduction Algorithm for Style Transfer Application
No full textA latent space represents data by embedding them in a multidimensional vector space. In this way an abstract estimation of any complex domain could be created. Empirical approach for exploring the latent space generated by known pre-trained model of human face images using a nonlinear dimensionality reduction algorithm is presented in this paper. One aim was to find more detailed entangled features (beard and hair color) between the real images and their representation, in artistic face portrait application. Experimental results showed that sparse vectors in the latent space could be useful to obtain optimal results with relatively low effort. To evaluate our work, we present the results of a survey that was sent to 25 thousand subscribers of the real world application and got around 360 responses. The main goal of the survey was to find some quantitative measurements that can be used in our research
Prime convolutional model: Breaking the ground for theoretical explainability
Full text linkIn this paper, we propose a new theoretical approach to Explainable AI. Following the Scientific Method, this approach consists of formulating, on the basis of empirical evidence, a mathematical model to explain and predict the behaviors of Neural Networks. We apply the method to a case study created in a controlled environment, which we call Prime Convolutional Model (p-Conv for short). p-Conv operates on a dataset consisting of the first one million natural numbers and is trained to identify the congruence classes modulo a given integer m. Its architecture uses a convolutional-type neural network that contextually processes a sequence of B consecutive numbers for each input. We take an empirical approach and exploit p-Conv to identify the congruence classes of numbers in a validation set using different values for m and B. The results show that the different behaviors of p-Conv (i.e., whether it can perform the task or not) can be modeled mathematically in terms of m and B. The inferred mathematical model reveals interesting patterns able to explain when and why p-Conv succeeds in performing task and, if not, which error pattern it follows
Investigating the Potential of Latent Space for the Classification of Paint Defects
Full text linkDefect detection methods have greatly assisted human operators in various fields, from textiles to surfaces and mechanical components, by facilitating decision-making processes and reducing visual fatigue. This area of research is widely recognized as a crossindustry concern, particularly in the manufacturing sector. Nevertheless, each specific application brings unique challenges that require tailored solutions. This paper presents
a novel framework for leveraging latent space representations in defect detection tasks, focusing on improving explainability while maintaining accuracy. This work delves into how latent spaces can be utilized by integrating unsupervised and supervised analyses. We propose a hybrid methodology that not only identifies known defects but also provides a mechanism for detecting anomalies and dynamically adapting to new defect
types. This dual approach supports human operators, reducing manual workload and enhancing interpretability
On the Construction of Numerical Models through a Prime Convolutional Approach
Full text linkIn this paper we apply neural network models to a set of natural numbers in order to classify the congruence classes modulo a given integer m ∈ {2, 3,..., 10}. We compare the performances of two kinds of architectures and of several input data representations. It turns out that these tasks are fully solved using a convolutional architecture and a special representation for the input data that exploits the prime factor decomposition of numbers
Multi-party Computation for Privacy and Security in Machine Learning: a Practical Review
Full text linkMachine Learning, particularly Deep Learning, is transforming society in any of its fundamental domains - healthcare, culture, finance, transportation, education, just to mention a few. However Machine Learning suffers from serious weaknesses in privacy and security due to the large amount of data
(datasets for training and parameters in trained models) and the probabilistic approximation inherent in any ML function. Multi-Party Computation (MPC) is a family of techniques and tactic with a sound scientific and operative base that can be applied to mitigate some relevant weaknesses of ML. In particular, privacy in training may be assured by MPC with federated learning techniques (these may be considered particular interpretations
and implementation of a general MPC method) and also security in training and inference may be enforced by continuous model testing using MPC is a technique that allows multiple parties to evaluate a machine learning model on their private data without revealing it to each other. This brief paper is a practical and essential review on how to use MPC to mitigate privacy and security issues in M
Toward Unbiased High-Quality Portraits through Latent-Space Evaluation
Full text linkImages, texts, voices, and signals can be synthesized by latent spaces in a multidimensional vector, which can be explored without the hurdles of noise or other interfering factors. In this paper, we present a practical use case that demonstrates the power of latent space in exploring complex realities such as image space. We focus on DaVinciFace, an AI-based system that explores the StyleGAN2 space to create a high-quality portrait for anyone in the style of the Renaissance genius Leonardo da Vinci. The user enters one of their portraits and receives the corresponding Da Vinci-style portrait as an output. Since most of Da Vinci’s artworks depict young and beautiful women (e.g., “La Belle Ferroniere”, “Beatrice de’ Benci”), we investigate the ability of DaVinciFace to account for other social categorizations, including gender, race, and age. The experimental results evaluate the effectiveness of our methodology on 1158 portraits acting on the vector representations of the latent space to produce high-quality portraits that retain the facial features of the subject’s social categories, and conclude that sparser vectors have a greater effect on these features. To objectively evaluate and quantify our results, we solicited human feedback via a crowd-sourcing campaign. Analysis of the human feedback showed a high tolerance for the loss of important identity features in the resulting portraits when the Da Vinci style is more pronounced, with some exceptions, including Africanized individuals
Going Beyond Counting First Authors in Author Co-citation Analysis
Full text linkThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
- …
