Christian Raul Salamea

Dr. Christian Raul Salamea

orcid

Research Focus & Metrics

Academic Career Length

20Years (2007-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
24 July 2019 in Applied Sciences
Verified
Journal Article
Open access
PDF file
David RomeroDavid Romero
David Romero
Christian Raul SalameaChristian Raul Salamea
Christian Raul Salamea
Closed-circuit television monitoring systems used for surveillance do not provide an immediate response in situations of danger such as armed robbery. In addition, they have multiple limitations when human operators perform the monitoring. For these reasons, a firearms detection system was developed using a new large database that was created from images extracted from surveillance videos of situations in which there are people with firearms. The system is made up of two parts—the “Front End” and “Back End”. The Front End is comprised of the YOLO object detection and localization system, and the Back End is made up of the firearms detection model that is developed in this work. These two systems are used to focus the detection system only in areas of the image where there are people, disregarding all other irrelevant areas. The performance of the firearm detection system was analyzed using multiple convolutional neural network (CNN) architectures, finding values up to 86% in metrics like recall and precision in a network configuration based on VGG Net using grayscale images.

12 December 2016 in IEEE Instrumentation & Measurement Magazine
Verified
Journal Article
Closed access
Rubén San-SegundoRubén San-Segundo
Rubén San-Segundo
JEJulián David Echeverry-Correa
Julián David Echeverry-Correa
Christian Raul SalameaChristian Raul Salamea
Christian Raul Salamea
JPJose Manuel Pardo
Jose Manuel Pardo
This paper presents an human sensing (HS) system based on Hidden Markov Models (HMMs) for classifying physical activities: walking, walking-upstairs, walking-downstairs, sitting, standing and lying down. The system includes a feature extractor (developed by the authors and presented in a previous work), an HMMs training module and an HAR module. All experiments have been done using a publicly available dataset named UCI Human Activity Recognition Using Smartphones. The final results using HMMs obtain comparable results to other recognition methods. Some improvements have been obtained when considering a discriminative HMM training procedure. The best result obtains an activity recognition error rate (ARER) of 2.5%. This work is focused on independent activity recognition and extends other works from the same authors focused on activity segmentation and feature extraction.

1 May 2014
Verified
Conference Paper
Open access
LDLuis Fernando D'Haro
Luis Fernando D'Haro
RCRicardo Cordoba
Ricardo Cordoba
Christian Raul SalameaChristian Raul Salamea
Christian Raul Salamea
JEJ. D. Echeverry
J. D. Echeverry
This paper presents new techniques with relevant improvements added to the primary system presented by our group to the Albayzin 2012 LRE competition, where the use of any additional corpora for training or optimizing the models was forbidden. In this work, we present the incorporation of an additional phonotactic subsystem based on the use of phone log-likelihood ratio features (PLLR) extracted from different phonotactic recognizers that contributes to improve the accuracy of the system in a 21.4% in terms of C avg (we also present results for the official metric during the evaluation, F act ). We will present how using these features at the phone state level provides significant improvements, when used together with dimensionality reduction techniques, especially PCA. We have also experimented with applying alternative SDC-like configurations on these PLLR features with additional improvements. Also, we will describe some modifications to the MFCC-based acoustic i-vector system which have also contributed to additional improvements. The final fused system outperformed the baseline in 27.4% in C avg .

21 June 2016
Verified
Conference Paper
Open access
Christian Raul SalameaChristian Raul Salamea
Christian Raul Salamea
LDLuis Fernando D'Haro
Luis Fernando D'Haro
RCRicardo Cordoba
Ricardo Cordoba
Rubén San-SegundoRubén San-Segundo
Rubén San-Segundo

10 March 2023
Verified
Conference Paper
Open access
Christian Raul SalameaChristian Raul Salamea
Christian Raul Salamea
Tarquino SánchezTarquino Sánchez
Tarquino Sánchez
XCXavier Alexander Calderón
Xavier Alexander Calderón
JGJavier Guaña-Moya
Javier Guaña-Moya
PCPaulo Castaneda-Romero
Paulo Castaneda-Romero
JRJessica Reina-Travez
Jessica Reina-Travez
The COVID-19 pandemic has changed the daily lives of all people worldwide, affecting not only society but also various sectors such as finance, tourism, etc. To counteract the pandemic, measures are required to detect contagions and take the necessary actions to prevent the virus spread. In this work, a Transfer Learning approach has been used to model COVID-19 coughs as a previous step to the diagnosis of the illness. The data set of the University of Cambridge, ComParE 2021 COVID-19 Cough Sub-Challenge, has been used, which consists of 725 samples of cough sounds from 397 people of which 119 have been diagnosed with positive COVID-19, besides, a data augmentation technique has been used to balance the data set. This work evaluates the performance of the pre-trained VGGish model for the classification of the audio cough signals as COVID or Not COVID cough. For this purpose, the VGGish model is used as a feature extractor and a convolutional neural network provides the final classification of the cough recordings to determine whether they are COVID-19 positive or negative. Despite the difficulty of the task, optimum results have been founded to detect negative cases obtaining up to 81% of precision. Considering the Unweighted Average Recall (UAR) as metric, the methodology proposed in this work has obtained an improvement up to 3% comparing to OpenSmile technique when the same database has been used.

20 March 2019 in Revista Iberoamericana de Automática e Informática Industrial
Verified
Journal Article
Open access
PDF file
Christian Raul SalameaChristian Raul Salamea
Christian Raul Salamea
SRSantiago Luna Romero
Santiago Luna Romero
En este trabajo se propone un filtrado adaptativo que incluye una etapa de calibración automática para la adquisición de señales EMG (Electromiografía). Se propone una técnica innovadora llamada “calibración automática” para minimizar el ruido provocado por el contacto de la piel con los sensores utilizados (electrodos) en condiciones de actividad física. Se utiliza filtrado adaptativo considerando que tanto la actividad física como la sudoración en personas son factores que alteran las condiciones de medición. La experimentación se ha realizado con personas que han desarrollado actividad física con diferentes condiciones de esfuerzo. Se ha utilizado la mejora relativa de la relación señal-ruido (RI-SNR) para comparar la técnica propuesta con filtros adaptativos que usan “ruido blanco” como señal de referencia. Este trabajo estuvo enfocado en los estimadores: Wiener, LMS y RLS, con mediciones realizadas antes y después de la actividad física. La técnica propuesta presenta una mejora de hasta un 45,49%, comparada con la correspondiente que utiliza “ruido blanco” para la calibración.