Mario Ignacio Chacon-Murguia
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Mario Ignacio Chacon-Murguia has a total of 62 co-authors. Here are the top co-authors by number of shared publications:
- Juan Alberto Ramirez-Quintana
Tecnologico Nacional de Mexico/I.T. Chihuahua
16 shared publications - Alma Delia Corral Saenz
Tecnologico Nacional de Mexico/I.T. Chihuahua
5 shared publications - Eduardo Rivas-Posada
Tecnologico Nacional de Mexico, Instituto Chihuahua, Chihuahua, Mexico
5 shared publications - Graciela Ramirez-Alonso
Universidad Autónoma de Chihuahua
4 shared publications - Abimael Guzman-Pando
Universidad Autónoma de Chihuahua
3 shared publications - Carlos Arzate-Quintana
Facultad de Medicina y Ciencias Biomédicas, Universidad Autónoma de Chihuahua, Chihuahua 31125, Mexico
3 shared publications
Latest Publications
- 8 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Latin America Transactions
- Vol. 23 (8) , 662-674
De-Occlusion Face Model based on Deep Occlusor Segmentation and Deep Inpainting Models
Image inpainting is a computer vision task that reconstructs missing image regions. Given its potential for various applications, it is an area of great interest. Despite advances in this field thanks to deep models such as autoencoders and generative adversarial networks, fundamental challenges persist, such as the causal interpretation of information loss and the risk of overfitting and lack of diversity in the features obtained with autoencoders. In this context, this article presents an innovative deep network model to solve occluded face inpainting. The model focuses on attributing the loss of information to the occlusion. The proposed model consists of two deep models: one for segmenting the object occluding the face, called SOCLNET, and another for reconstructing the face, IFACENET. SOCLNET is an improvement of the DeepLabv3 network by adding self-attention mechanisms. IFACENET is based on an autoencoder with an ensemble learning approach in the encoder to improve the diversity of the extracted features. SOCLNET was evaluated to demonstrate that the segmentation of occluding objects works adequately, even on out-of-distribution images. Its performance metrics were Pixel Accuracy = 0.93 and IoU = 0.788. The IFACENET model was compared against other state-of-the-art models using the Celeb-HQ database. The quantitative results of IFACENET show an average performance of SSIM = 0.95, PSNR = 26.813, and L1 = 0.261 with different mask values, being competitive with the state of the art. Additionally, qualitative results of IFACENET are shown to demonstrate the visual outcomes of face inpainting. Based on those results, it can be concluded that the proposed model effectively solves the reconstruction of occluded faces, opening new perspectives in the research of image reconstruction.
- 8 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Latin America Transactions
- Vol. 23 (8) , 662-674
De-Occlusion Face Model based on Deep Occlusor Segmentation and Deep Inpainting Models
Image inpainting is a computer vision task that reconstructs missing image regions. Given its potential for various applications, it is an area of great interest. Despite advances in this field thanks to deep models such as autoencoders and generative adversarial networks, fundamental challenges persist, such as the causal interpretation of information loss and the risk of overfitting and lack of diversity in the features obtained with autoencoders. In this context, this article presents an innovative deep network model to solve occluded face inpainting. The model focuses on attributing the loss of information to the occlusion. The proposed model consists of two deep models: one for segmenting the object occluding the face, called SOCLNET, and another for reconstructing the face, IFACENET. SOCLNET is an improvement of the DeepLabv3 network by adding self-attention mechanisms. IFACENET is based on an autoencoder with an ensemble learning approach in the encoder to improve the diversity of the extracted features. SOCLNET was evaluated to demonstrate that the segmentation of occluding objects works adequately, even on out-of-distribution images. Its performance metrics were Pixel Accuracy = 0.93 and IoU = 0.788. The IFACENET model was compared against other state-of-the-art models using the Celeb-HQ database. The quantitative results of IFACENET show an average performance of SSIM = 0.95, PSNR = 26.813, and L1 = 0.261 with different mask values, being competitive with the state of the art. Additionally, qualitative results of IFACENET are shown to demonstrate the visual outcomes of face inpainting. Based on those results, it can be concluded that the proposed model effectively solves the reconstruction of occluded faces, opening new perspectives in the research of image reconstruction.
- 18 June 2025
- Published by Springer Nature
- p. 144-153
Deep Neural Networks and Log-Mel Spectrogram for Emotion Recognition Through Spanish Speech
- 18 June 2025
- Published by Springer Nature
- p. 144-153
Deep Neural Networks and Log-Mel Spectrogram for Emotion Recognition Through Spanish Speech
- Open Access
- 9 February 2025
- Published by MDPI AG in Big Data and Cognitive Computing
- Vol. 9 (2) , 36
Novel Extreme-Lightweight Fully Convolutional Network for Low Computational Cost in Microbiological and Cell Analysis: Detection, Quantification, and Segmentation
Integrating deep learning into microbiological and cell analysis from microscopic image samples has gained significant attention in recent years, driven by the rise of novel medical technologies and pressing global health challenges. Numerous methods for segmentation and classification in microscopic images have emerged in the literature. However, key challenges persist due to the limited development of specialized deep learning models to accurately detect and quantify microorganisms and cells from microscopic samples. In response to this gap, this paper introduces MBnet, an Extreme-Lightweight Neural Network for Microbiological and Cell Analysis. MBnet is a binary segmentation method based on a Fully Convolutional Network designed to detect and quantify microorganisms and cells, featuring a low computational cost architecture with only 575 parameters. Its innovative design includes a foreground module and an encoder–decoder structure composed of traditional, depthwise, and separable convolution layers. These layers integrate color, orientation, and morphological features to generate an understanding of different contexts in microscopic sample images for binary segmentation. Experiments were conducted using datasets containing bacteria, yeast, and blood cells. The results suggest that MBnet outperforms other popular networks in the literature in counting, detecting, and segmenting cells and unicellular microorganisms. These findings underscore the potential of MBnet as a highly efficient solution for real-world applications in health monitoring and bioinformatics.
- 9 February 2025
- Published by MDPI AG in Big Data and Cognitive Computing
- Vol. 9 (2) , 36
Open AccessNovel Extreme-Lightweight Fully Convolutional Network for Low Computational Cost in Microbiological and Cell Analysis: Detection, Quantification, and Segmentation
Integrating deep learning into microbiological and cell analysis from microscopic image samples has gained significant attention in recent years, driven by the rise of novel medical technologies and pressing global health challenges. Numerous methods for segmentation and classification in microscopic images have emerged in the literature. However, key challenges persist due to the limited development of specialized deep learning models to accurately detect and quantify microorganisms and cells from microscopic samples. In response to this gap, this paper introduces MBnet, an Extreme-Lightweight Neural Network for Microbiological and Cell Analysis. MBnet is a binary segmentation method based on a Fully Convolutional Network designed to detect and quantify microorganisms and cells, featuring a low computational cost architecture with only 575 parameters. Its innovative design includes a foreground module and an encoder–decoder structure composed of traditional, depthwise, and separable convolution layers. These layers integrate color, orientation, and morphological features to generate an understanding of different contexts in microscopic sample images for binary segmentation. Experiments were conducted using datasets containing bacteria, yeast, and blood cells. The results suggest that MBnet outperforms other popular networks in the literature in counting, detecting, and segmenting cells and unicellular microorganisms. These findings underscore the potential of MBnet as a highly efficient solution for real-world applications in health monitoring and bioinformatics.
- 17 October 2024
- Published by Springer Nature
- p. 113-123
Emotion Recognition Method Based on EEG Signal Processing, Simplified Inception Network and Discrete Model
- 17 October 2024
- Published by Springer Nature
- p. 113-123
Emotion Recognition Method Based on EEG Signal Processing, Simplified Inception Network and Discrete Model
- Open Access
- 12 October 2024
- Published by Tecnológico Nacional de México / Instituto Tecnológico de Chihuahua in Revista ELECTRO
- Vol. 46 (1) , 351-356
Reconocimiento de Emociones Mediante Imágenes de Expresiones Faciales Basado en Arquitectura de Red Siamesa
El interés por estado emocional de una persona se ha incrementado en gran medida en el sector salud debido a las consecuencias que puede tener una situación de un estado emocional deteriorado no atendido. Por lo tanto, en este trabajo se propone un modelo de red profunda con arquitectura siamesa para el reconocimiento de emociones humanas por medio de imágenes, y que sea capaz de funcionar sobre un sistema embebido de manera rápida, portable y sin complicaciones de incompatibilidad de paquetes o una pobre optimización. Para desarrollar el modelo final de red siamesa se realizó una etapa previa llevando a cabo entrenamientos con pares de emociones que siguieran estas características: que fueran similares entre sí, que fueran medianamente similares, y que fueran distintas. El número de emociones se fue incrementando hasta que el modelo mantuviera una precisión general por arriba del 70%. Esta restricción se logró alcanzar con 5 emociones. El modelo final resultó tener un desempeño de 83.83% de precisión.
- 12 October 2024
- Published by Tecnológico Nacional de México / Instituto Tecnológico de Chihuahua in Revista ELECTRO
- Vol. 46 (1) , 351-356
Open AccessReconocimiento de Emociones Mediante Imágenes de Expresiones Faciales Basado en Arquitectura de Red Siamesa
El interés por estado emocional de una persona se ha incrementado en gran medida en el sector salud debido a las consecuencias que puede tener una situación de un estado emocional deteriorado no atendido. Por lo tanto, en este trabajo se propone un modelo de red profunda con arquitectura siamesa para el reconocimiento de emociones humanas por medio de imágenes, y que sea capaz de funcionar sobre un sistema embebido de manera rápida, portable y sin complicaciones de incompatibilidad de paquetes o una pobre optimización. Para desarrollar el modelo final de red siamesa se realizó una etapa previa llevando a cabo entrenamientos con pares de emociones que siguieran estas características: que fueran similares entre sí, que fueran medianamente similares, y que fueran distintas. El número de emociones se fue incrementando hasta que el modelo mantuviera una precisión general por arriba del 70%. Esta restricción se logró alcanzar con 5 emociones. El modelo final resultó tener un desempeño de 83.83% de precisión.
- 22 February 2024
- Published by Springer Nature in Neural Computing and Applications
- Vol. 36 (16) , 9073-9094
Meta-learning methodology based on meta-unsupervised algorithm for meta-model selection to solve few-shot base-tasks
- 22 February 2024
- Published by Springer Nature in Neural Computing and Applications
- Vol. 36 (16) , 9073-9094
Meta-learning methodology based on meta-unsupervised algorithm for meta-model selection to solve few-shot base-tasks
- 29 June 2023
- Published by Springer Nature in Soft Computing
- Vol. 28 (15-16) , 8561-8577
AENCIC: a method to estimate the number of clusters based on image complexity to be used in fuzzy clustering algorithms for image segmentation
- 29 June 2023
- Published by Springer Nature in Soft Computing
- Vol. 28 (15-16) , 8561-8577
AENCIC: a method to estimate the number of clusters based on image complexity to be used in fuzzy clustering algorithms for image segmentation
- 28 June 2023
- Published by Elsevier in Computers in Biology and Medicine
- Vol. 163, 107200
Automatic base-model selection for white blood cell image classification using meta-learning
- 28 June 2023
- Published by Elsevier in Computers in Biology and Medicine
- Vol. 163, 107200
Automatic base-model selection for white blood cell image classification using meta-learning
- Open Access
- 11 June 2023
- Published by Universidad de Guadalajara in RECIBE, REVISTA ELECTRÓNICA DE COMPUTACIÓN, INFORMÁTICA, BIOMÉDICA Y ELECTRÓNICA
- Vol. 12 (1) , C6-43
Revisión de Problemas en la Detección de Objetos en Imágenes y Videos Digitales
En las últimas décadas, la detección de objetos ha sido una tarea muy importante en el área de visión por computadora, ya que la detección de objetos localiza y clasifica uno o más objetos en una imagen o videos. En este artículo, se presenta una revisión de artículos y se describen técnicas clásicas y de aprendizaje profundas utilizadas para la detección de objetos. Además, se realiza una revisión de trabajos recientes sobre la detección de objetos, enfocándose en cómo se solucionan algunos de sus problemas más relevantes. Los problemas que se abarcan son: oclusión, confusión, información contextual, cambios en la iluminación, objetos pequeños y cambios de escala, variación entre la misma clase y diferentes clases, y deformación y cambios de pose. Se espera que este artículo sirva para que los interesados en el área conozcan ideas y enfoques para resolver problemas existentes en la detección de objetos incluyendo los últimos avances.
- 11 June 2023
- Published by Universidad de Guadalajara in RECIBE, REVISTA ELECTRÓNICA DE COMPUTACIÓN, INFORMÁTICA, BIOMÉDICA Y ELECTRÓNICA
- Vol. 12 (1) , C6-43
Open AccessRevisión de Problemas en la Detección de Objetos en Imágenes y Videos Digitales
En las últimas décadas, la detección de objetos ha sido una tarea muy importante en el área de visión por computadora, ya que la detección de objetos localiza y clasifica uno o más objetos en una imagen o videos. En este artículo, se presenta una revisión de artículos y se describen técnicas clásicas y de aprendizaje profundas utilizadas para la detección de objetos. Además, se realiza una revisión de trabajos recientes sobre la detección de objetos, enfocándose en cómo se solucionan algunos de sus problemas más relevantes. Los problemas que se abarcan son: oclusión, confusión, información contextual, cambios en la iluminación, objetos pequeños y cambios de escala, variación entre la misma clase y diferentes clases, y deformación y cambios de pose. Se espera que este artículo sirva para que los interesados en el área conozcan ideas y enfoques para resolver problemas existentes en la detección de objetos incluyendo los últimos avances.
- 9 June 2023
- Published by Springer Nature
- p. 221-230
Leukocyte Detection with Novel Fully Convolutional Network and a New Dataset of Blood Smear Complete Samples
- 9 June 2023
- Published by Springer Nature
- p. 221-230
Leukocyte Detection with Novel Fully Convolutional Network and a New Dataset of Blood Smear Complete Samples