arXiv subjects

Miguel Carrasco

Publications and source records attributed to Miguel Carrasco.

9 recordsLinked to original sources

arXiv2026

Macroscopic bioinspired magnetic active matter and the physical limits of magnetotaxis

Magnetotactic bacteria (MTB) are endowed with an exquisite orientation mechanism allowing them to swim along the geomagnetic field lines. This mechanism consists of a chain of bio-synthesized magnetic nano-crystals that endow MTB with a permanent magnetic moment. Although the physics behind the minimum size of this biological compass is well understood, it is yet unclear what sets its maximum size. Here we combine macroscopic bioinspired experiments, calibrated simulations, and analytic estimates to show that increasing dipolar strength can drive magnetic active matter from a freely swimming regime into clustered states. Using a physical model with parameters relevant to MTB, we infer a plausible physical upper bound on useful magnetosome-chain magnetic moments: beyond a threshold, clustering and the formation of compound bodies are expected to hinder effective swimming and reduce magnetotactic performance. Our macroscopic bio-inspired experiment and physical model show how long-range magnetic interactions reshape the phase behavior of anisotropic active matter and provide a programmable platform for studying magnetic active matter across scales.

cond-mat.soft↗

arXiv2025

Testing chatbots on the creation of encoders for audio conditioned image generation

On one hand, recent advances in chatbots has led to a rising popularity in using these models for coding tasks. On the other hand, modern generative image models primarily rely on text encoders to translate semantic concepts into visual representations, even when there is clear evidence that audio can be employed as input as well. Given the previous, in this work, we explore whether state-of-the-art conversational agents can design effective audio encoders to replace the CLIP text encoder from Stable Diffusion 1.5, enabling image synthesis directly from sound. We prompted five publicly available chatbots to propose neural architectures to work as these audio encoders, with a set of well-explained shared conditions. Each valid suggested encoder was trained on over two million context related audio-image-text observations, and evaluated on held-out validation and test sets using various metrics, together with a qualitative analysis of their generated images. Although almost all chatbots generated valid model designs, none achieved satisfactory results, indicating that their audio embeddings failed to align reliably with those of the original text encoder. Among the proposals, the Gemini audio encoder showed the best quantitative metrics, while the Grok audio encoder produced more coherent images (particularly, when paired with the text encoder). Our findings reveal a shared architectural bias across chatbots and underscore the remaining coding gap that needs to be bridged in future versions of these models. We also created a public demo so everyone could study and try out these audio encoders. Finally, we propose research questions that should be tackled in the future, and encourage other researchers to perform more focused and highly specialized tasks like this one, so the respective chatbots cannot make use of well-known solutions and their creativity/reasoning is fully tested.

cs.SD↗

arXiv2025

Effectively obtaining acoustic, visual and textual data from videos

The increasing use of machine learning models has amplified the demand for high-quality, large-scale multimodal datasets. However, the availability of such datasets, especially those combining acoustic, visual and textual data, remains limited. This paper addresses this gap by proposing a method to extract related audio-image-text observations from videos. We detail the process of selecting suitable videos, extracting relevant data pairs, and generating descriptive texts using image-to-text models. Our approach ensures a robust semantic connection between modalities, enhancing the utility of the created datasets for various applications. We also discuss the challenges encountered and propose solutions to improve data quality. The resulting datasets, publicly available, aim to support and advance research in multimodal data analysis and machine learning.

cs.MM↗

arXiv2025

Vision Transformer attention alignment with human visual perception in aesthetic object evaluation

Visual attention mechanisms play a crucial role in human perception and aesthetic evaluation. Recent advances in Vision Transformers (ViTs) have demonstrated remarkable capabilities in computer vision tasks, yet their alignment with human visual attention patterns remains underexplored, particularly in aesthetic contexts. This study investigates the correlation between human visual attention and ViT attention mechanisms when evaluating handcrafted objects. We conducted an eye-tracking experiment with 30 participants (9 female, 21 male, mean age 24.6 years) who viewed 20 artisanal objects comprising basketry bags and ginger jars. Using a Pupil Labs eye-tracker, we recorded gaze patterns and generated heat maps representing human visual attention. Simultaneously, we analyzed the same objects using a pre-trained ViT model with DINO (Self-DIstillation with NO Labels), extracting attention maps from each of the 12 attention heads. We compared human and ViT attention distributions using Kullback-Leibler divergence across varying Gaussian parameters (sigma=0.1 to 3.0). Statistical analysis revealed optimal correlation at sigma=2.4 +-0.03, with attention head #12 showing the strongest alignment with human visual patterns. Significant differences were found between attention heads, with heads #7 and #9 demonstrating the greatest divergence from human attention (p< 0.05, Tukey HSD test). Results indicate that while ViTs exhibit more global attention patterns compared to human focal attention, certain attention heads can approximate human visual behavior, particularly for specific object features like buckles in basketry items. These findings suggest potential applications of ViT attention mechanisms in product design and aesthetic evaluation, while highlighting fundamental differences in attention strategies between human perception and current AI models.

cs.CV↗

arXiv2024

Level of agreement between emotions generated by Artificial Intelligence and human evaluation: a methodological proposal

Images are capable of conveying emotions, but emotional experience is highly subjective. Advances in artificial intelligence have enabled the generation of images based on emotional descriptions. However, the level of agreement between the generative images and human emotional responses has not yet been evaluated. To address this, 20 artistic landscapes were generated using StyleGAN2-ADA. Four variants evoking positive emotions (contentment, amusement) and negative emotions (fear, sadness) were created for each image, resulting in 80 pictures. An online questionnaire was designed using this material, in which 61 observers classified the generated images. Statistical analyses were performed on the collected data to determine the level of agreement among participants, between the observer's responses, and the AI-generated emotions. A generally good level of agreement was found, with better results for negative emotions. However, the study confirms the subjectivity inherent in emotional evaluation.

cs.CV↗

arXiv2023

OWAdapt: An adaptive loss function for deep learning using OWA operators

In this paper, we propose a fuzzy adaptive loss function for enhancing deep learning performance in classification tasks. Specifically, we redefine the cross-entropy loss to effectively address class-level noise conditions, including the challenging problem of class imbalance. Our approach introduces aggregation operators, leveraging the power of fuzzy logic to improve classification accuracy. The rationale behind our proposed method lies in the iterative up-weighting of class-level components within the loss function, focusing on those with larger errors. To achieve this, we employ the ordered weighted average (OWA) operator and combine it with an adaptive scheme for gradient-based learning. Through extensive experimentation, our method outperforms other commonly used loss functions, such as the standard cross-entropy or focal loss, across various binary and multiclass classification tasks. Furthermore, we explore the influence of hyperparameters associated with the OWA operators and present a default configuration that performs well across different experimental settings.

cs.LG↗

arXiv2022

Analysis of the use of color and its emotional relationship in visual creations based on experiences during the context of the COVID-19 pandemic

Color is a complex communicative element that helps us understand and evaluate our environment. At the level of artistic creation, this component influences both the formal aspects of the composition and the symbolic weight, directly affecting the construction and transmission of the message that you want to communicate, creating a specific emotional reaction. During the COVID-19 pandemic, people generated countless images transmitting this event's subjective experiences. Using the repository of images created in the Instagram account CAM (The COVID Art Museum), we propose a methodology to understand the use of color and its emotional relationship in this context. The process considers two stages in parallel that are then combined. First, emotions are extracted and classified from the CAM dataset images through a convolutional neural network. Second, we extract the colors and their harmonies through a clustering process. Once both processes are completed, we combine the results generating an expanded discussion on the usage of color, harmonies, and emotion. The results indicate that warm colors are prevalent in the sample, with a preference for analog compositions over complementary ones. The relationship between emotions and these compositions shows a trend in positive emotions, reinforced by the results of the algorithm a priori and the emotional relationship analysis of the attributes of color (hue, chroma, and lighting).

cs.CV↗

arXiv2021

Quantitative analysis of visual representation of sign elements in COVID-19 context

Representation is the way in which human beings re-present the reality of what is happening, both externally and internally. Thus, visual representation as a means of communication uses elements to build a narrative, just as spoken and written language do. We propose using computer analysis to perform a quantitative analysis of the elements used in the visual creations that have been produced in reference to the epidemic, using the images compiled in The Covid Art Museum's Instagram account to analyze the different elements used to represent subjective experiences with regard to a global event. This process has been carried out with techniques based on machine learning to detect objects in the images so that the algorithm can be capable of learning and detecting the objects contained in each study image. This research reveals that the elements that are repeated in images to create narratives and the relations of association that are established in the sample, concluding that, despite the subjectivity that all creation entails, there are certain parameters of shared and reduced decisions when it comes to selecting objects to be included in visual representations

cs.CV↗

arXiv2020

Stable and unstable trajectories in a dipolar chain

In classical mechanics, solutions can be classified according to their stability. Each of them is part of the possible trajectories of the system. However, the signatures of unstable solutions are hard to observe in an experiment, and most of the times if the experimental realization is adiabatic, they are considered just a nuisance. Here we use a small number of XY magnetic dipoles subject to an external magnetic field for studying the origin of their collective magnetic response. Using bifurcation theory we have found all the possible solutions being stable or unstable, and explored how those solutions are naturally connected by points where the symmetries of the system are lost or restored. Unstable solutions that reveal the symmetries of the system are found to be the culprit that shape hysteresis loops in this system. The complexity of the solutions for the nonlinear dynamics is analyzed using the concept of boundary basin entropy, finding that the damping time scale is critical for the emergence of fractal structures in the basins of attraction. Furthermore, we numerically found domain wall solutions that are the smallest possible realizations of transverse walls and vortex walls in magnetism. We experimentally confirmed their existence and stability showing that our system is a suitable platform to study domain wall dynamics at the macroscale.

nlin.PS↗