Open Access
Enhancing DreamBooth with LoRA for generating unlimited characters with stable diffusion
(IEEE, 2024-09-09) Pascual Casas, Rubén; Maiza Coupin, Adrián Mikel; Sesma Sara, Mikel; Paternain Dallo, Daniel; Galar Idoate, Mikel; Estadística, Informática y Matemáticas; Estatistika, Informatika eta Matematika; Institute of Smart Cities - ISC; Universidad Pública de Navarra / Nafarroako Unibertsitate Publikoa, PJUPNA2023-11377
This paper addresses the challenge of generating unlimited new and distinct characters that encompass the style and shared visual characteristics of a limited set of human designed characters. This is a relevant problem in the audiovisual industry, as the ability to rapidly produce original characters that adhere to specific characteristics greatly increases the possibilities in the production of movies, series, or video games. Our solution is built upon DreamBooth, a widely extended fine-tuning method for text-to-image models. We propose an adaptation focusing on two main challenges: the impracticality of relying on detailed image prompts for character description and the few-shot learning scenario with a limited set of characters available for training. To solve these issues, we introduce additional character-specific tokens to DreamBooth training and remove its class-specific regularization dataset. For an unlimited generation of characters, we propose the usage of random tokens and random embeddings. This proposal is tested on two specialized datasets and the results shows our method¿s capability to produce diverse characters that adhere to a style and visual characteristics. An ablation study to analyze the contributions of the proposed modifications is also developed.
Open Access
Less can be more: representational vs. stereotypical gender bias in facial expression recognition
(Springer, 2024-10-14) Domínguez Catena, Iris; Paternain Dallo, Daniel; Jurío Munárriz, Aránzazu; Galar Idoate, Mikel; Estadística, Informática y Matemáticas; Estatistika, Informatika eta Matematika; Universidad Publica de Navarra / Nafarroako Unibertsitate Publikoa
Machine learning models can inherit biases from their training data, leading to discriminatory or inaccurate predictions. This is particularly concerning with the increasing use of large, unsupervised datasets for training foundational models. Traditionally, demographic biases within these datasets have not been well-understood, limiting our ability to understand how they propagate to the models themselves. To address this issue, this paper investigates the propagation of demographic biases from datasets into machine learning models. We focus on the gender demographic component, analyzing two types of bias: representational and stereotypical. For our analysis, we consider the domain of facial expression recognition (FER), a field known to exhibit biases in most popular datasets. We use Affectnet, one of the largest FER datasets, as our baseline for carefully designing and generating subsets that incorporate varying strengths of both representational and stereotypical bias. Subsequently, we train several models on these biased subsets, evaluating their performance on a common test set to assess the propagation of bias into the models¿ predictions. Our results show that representational bias has a weaker impact than expected. Models exhibit a good generalization ability even in the absence of one gender in the training dataset. Conversely, stereotypical bias has a significantly stronger impact, primarily concentrated on the biased class, although it can also influence predictions for unbiased classes. These results highlight the need for a bias analysis that differentiates between types of bias, which is crucial for the development of effective bias mitigation strategies.
Open Access
Generación ilimitada de personajes mediante Stable Diffusion con DreamBooth y LoRA
(CAEPIA, 2024) Pascual Casas, Rubén; Maiza Coupin, Adrián Mikel; Sesma Sara, Mikel; Paternain Dallo, Daniel; Galar Idoate, Mikel; Estadística, Informática y Matemáticas; Estatistika, Informatika eta Matematika; Institute of Smart Cities - ISC; Universidad Pública de Navarra / Nafarroako Unibertsitate Publikoa, PJUPNA2023-11377; Gobierno de Navarra / Nafarroako Gobernua
Este artículo aborda el reto de generar un número ilimitado de personajes nuevos, y distintos, que engloben el estilo y las características visuales compartidas de un conjunto limitado de personajes diseñados por un humano. Este es un problema de gran relevancia en la industria audiovisual, ya que la capacidad de producir rápidamente personajes originales que se adhieran a unas características específicas aumenta enormemente las posibilidades en la producción de películas, series o videojuegos. Nuestra solución se basa en DreamBooth, un método de ajuste de modelos generativos de texto a imagen ampliamente extendido. Proponemos una adaptación centrada en dos retos principales: lo poco práctico que resulta utilizar prompts detallados de las imágenes para describir los personajes y la complejidad del ajuste de modelos a partir de un conjunto limitado de personajes. Para resolver estos problemas, introducimos en el entrenamiento de DreamBooth tokens adicionales específicos para cada personaje y eliminamos el conjunto de datos de regularización. Para generar personajes de manera ilimitada, proponemos el uso de tokens y embeddings aleatorios. Comprobamos la utilidad de la propuesta utilizando dos conjuntos de datos diferentes. Los resultados obtenidos muestran la capacidad de nuestro método para producir personajes diversos que se adhieren a un estilo y a unas características visuales concretas. Finalmente, desarrollamos un estudio de ablación.
Open Access
DSAP: analyzing bias through demographic comparison of datasets
(Elsevier, 2024-10-29) Domínguez Catena, Iris; Paternain Dallo, Daniel; Galar Idoate, Mikel; Estadística, Informática y Matemáticas; Estatistika, Informatika eta Matematika; Universidad Publica de Navarra / Nafarroako Unibertsitate Publikoa ; Gobierno de Navarra / Nafarroako Gobernua
In the last few years, Artificial Intelligence (AI) systems have become increasingly widespread. Unfortunately, these systems can share many biases with human decision-making, including demographic biases. Often, these biases can be traced back to the data used for training, where large uncurated datasets have become the norm. Despite our awareness of these biases, we still lack general tools to detect, quantify, and compare them across different datasets. In this work, we propose DSAP (Demographic Similarity from Auxiliary Profiles), a two-step methodology for comparing the demographic composition of datasets. First, DSAP uses existing demographic estimation models to extract a dataset's demographic profile. Second, it applies a similarity metric to compare the demographic profiles of different datasets. While these individual components are well-known, their joint use for demographic dataset comparison is novel and has not been previously addressed in the literature. This approach allows three key applications: the identification of demographic blind spots and bias issues across datasets, the measurement of demographic bias, and the assessment of demographic shifts over time. DSAP can be used on datasets with or without explicit demographic information, provided that demographic information can be derived from the samples using auxiliary models, such as those for image or voice datasets. To show the usefulness of the proposed methodology, we consider the Facial Expression Recognition task, where demographic bias has previously been found. The three applications are studied over a set of twenty datasets with varying properties. The code is available at https://github.com/irisdominguez/DSAP.
Open Access
d-Choquet integrals: Choquet integrals based on dissimilarities
(Elsevier, 2020) Bustince Sola, Humberto; Mesiar, Radko; Fernández Fernández, Francisco Javier; Galar Idoate, Mikel; Paternain Dallo, Daniel; Altalhi, A. H.; Pereira Dimuro, Graçaliz; Bedregal, Benjamin; Takáč, Zdenko; Estatistika, Informatika eta Matematika; Institute of Smart Cities - ISC; Estadística, Informática y Matemáticas; Universidad Pública de Navarra / Nafarroako Unibertsitate Publikoa, PJUPNA13
The paper introduces a new class of functions from [0,1]n to [0,n] called d-Choquet integrals. These functions are a generalization of the 'standard' Choquet integral obtained by replacing the difference in the definition of the usual Choquet integral by a dissimilarity function. In particular, the class of all d-Choquet integrals encompasses the class of all 'standard' Choquet integrals but the use of dissimilarities provides higher flexibility and generality. We show that some d-Choquet integrals are aggregation/pre-aggregation/averaging/functions and some of them are not. The conditions under which this happens are stated and other properties of the d-Choquet integrals are studied.
Open Access
Discrete IV dG-Choquet integrals with respect to admissible orders
(Elsevier, 2021) Takáč, Zdenko; Uriz Martín, Mikel Xabier; Galar Idoate, Mikel; Paternain Dallo, Daniel; Bustince Sola, Humberto; Estatistika, Informatika eta Matematika; Ingeniaritza Elektrikoa, Elektronikoaren eta Telekomunikazio Ingeniaritzaren; Institute of Smart Cities - ISC; Estadística, Informática y Matemáticas; Ingeniería Eléctrica, Electrónica y de Comunicación; Universidad Pública de Navarra / Nafarroako Unibertsitate Publikoa
In this work, we introduce the notion of dG-Choquet integral, which generalizes the discrete Choquet integral replacing, in the first place, the difference between inputs represented by closed subintervals of the unit interval [0,1] by a dissimilarity function; and we also replace the sum by more general appropriate functions. We show that particular cases of dG-Choquet integral are both the discrete Choquet integral and the d-Choquet integral. We define interval-valued fuzzy measures and we show how they can be used with dG-Choquet integrals to define an interval-valued discrete Choquet integral which is monotone with respect to admissible orders. We finally study the validity of this interval-valued Choquet integral by means of an illustrative example in a classification problem. © 2021
Open Access
Dissimilarity based choquet integrals
(Springer, 2020) Bustince Sola, Humberto; Mesiar, Radko; Fernández Fernández, Francisco Javier; Galar Idoate, Mikel; Paternain Dallo, Daniel; Estadística, Informática y Matemáticas; Estatistika, Informatika eta Matematika
In this paper, in order to generalize the Choquet integral, we replace the difference between inputs in its definition by a restricted dissimilarity function and refer to the obtained function as d-Choquet integral. For some particular restricted dissimilarity function the corresponding d-Choquet integral with respect to a fuzzy measure is just the ‘standard’ Choquet integral with respect to the same fuzzy measure. Hence, the class of all d-Choquet integrals encompasses the class of all 'standard' Choquet integrals. This approach allows us to construct a wide class of new functions, d-Choquet integrals, that are possibly, unlike the 'standard' Choquet integral, outside of the scope of aggregation functions since the monotonicity is, for some restricted dissimilarity function, violated and also the range of such functions can be wider than [0, 1], in particular it can be [0, n].
Open Access
On the influence of admissible orders in IVOVO
(Springer, 2019) Uriz Martín, Mikel Xabier; Paternain Dallo, Daniel; Bustince Sola, Humberto; Galar Idoate, Mikel; Estatistika, Informatika eta Matematika; Institute of Smart Cities - ISC; Estadística, Informática y Matemáticas; Universidad Pública de Navarra / Nafarroako Unibertsitate Publikoa, PJUPNA13
It is known that when dealing with interval-valued data, there exist problems associated with the non-existence of a total order. In this work we investigate a reformulation of an interval-valued decomposition strategy for multi-class problems called IVOVO, and we analyze the effectiveness of considering different admissible orders in the aggregation phase of IVOVO. We demonstrate that the choice of an appropriate admissible order allows the method to obtain significant differences in terms of accuracy.
Open Access
A survey of fingerprint classification Part II: experimental analysis and ensemble proposal
(Elsevier, 2015) Galar Idoate, Mikel; Derrac, Joaquín; Peralta, Daniel; Triguero, Isaac; Paternain Dallo, Daniel; López Molina, Carlos; García, Salvador; Benítez, José Manuel; Pagola Barrio, Miguel; Barrenechea Tartas, Edurne; Bustince Sola, Humberto; Herrera, Francisco; Automática y Computación; Automatika eta Konputazioa
In the first part of this paper we reviewed the fingerprint classification literature from two different perspectives: the feature extraction and the classifier learning. Aiming at answering the question of which among the reviewed methods would perform better in a real implementation we end up in a discussion which showed the difficulty in answering this question. No previous comparison exists in the literature and comparisons among papers are done with different experimental frameworks. Moreover, the difficulty in implementing published methods was stated due to the lack of details in their description, parameters and the fact that no source code is shared. For this reason, in this paper we will go through a deep experimental study following the proposed double perspective. In order to do so, we have carefully implemented some of the most relevant feature extraction methods according to the explanations found in the corresponding papers and we have tested their performance with different classifiers, including those specific proposals made by the authors. Our aim is to develop an objective experimental study in a common framework, which has not been done before and which can serve as a baseline for future works on the topic. This way, we will not only test their quality, but their reusability by other researchers and will be able to indicate which proposals could be considered for future developments. Furthermore, we will show that combining different feature extraction models in an ensemble can lead to a superior performance, significantly increasing the results obtained by individual models.
Open Access
An empirical study on supervised and unsupervised fuzzy measure construction methods in highly imbalanced classification
(IEEE, 2020) Uriz Martín, Mikel Xabier; Paternain Dallo, Daniel; Bustince Sola, Humberto; Galar Idoate, Mikel; Estatistika, Informatika eta Matematika; Institute of Smart Cities - ISC; Estadística, Informática y Matemáticas; Universidad Pública de Navarra / Nafarroako Unibertsitate Publikoa
The design of an ensemble of classifiers involves the definition of an aggregation mechanism that produces a single response obtained from the information provided by the classifiers. A specific aggregation methodology that has been studied in the literature is the use of fuzzy integrals, such as the Choquet or the Sugeno integral, where the associated fuzzy measure tries to represent the interaction existing between the classifiers of the ensemble. However, defining the big number of coefficients of a fuzzy measure is not a trivial task and therefore, many different algorithms have been proposed. These can be split into supervised and unsupervised, each class having different learning mechanisms and particularities. Since there is no clear knowledge about the correct method to be used, in this work we propose an experimental study for comparing the performance of eight different learning algorithms under the same framework of imbalanced dataset. Moreover, we also compare the specific fuzzy integral (Choquet or Sugeno) and their synergies with the different fuzzy measure construction methods.