There are about 33,000 species of fish found in underwater environments, each of which is distinguished by distinct visual traits such as shape, color, and head shape, these traits are complex for the average person to recognize. Scientists and aquaculturists use photography and video cameras to quantify and determine the state and shape of the different species. Providing tools like computer vision algorithms using convolutional neuronal networks (CNN). This chapter proposes a dataset of images of three distinct fish species: Goldfish (Carassius auratus), Molly (Poecilia sphenops), and Zebra (Danio rerio), each with a unique number of images. The data were collected using a camera in a natural environment where the species were housed in an aquaculturist-protected environment. The dataset made up of more than 300 images of each species will be trained in different CNN for species identification applications. The models are trained with images with several pre-processing strategies applied previously. The performance metrics obtained from the proposed dataset are validated and tested using CNN such as YOLOV5, and residual neural networks such as RESNET. The model YOLOV5 achieved a precision of 85 percent and RESNET 92 percent, these architectures are evaluated on edge computing devices such Jetson Nano, and Jetson TX2. <!----> <!----> <!----> <!---->

Due to the contingency situation that has been generated in various parts of the world and the declaration of a pandemic carried out by the World Health Organization against the Sars-CoV-2 virus, various people, educational institutions and companies are carrying out the development of mechanical ventilators that can meet the need for this equipment in their countries. This paper shows experiences obtained in the design and construction of a transitional mechanical ventilator that allows compliance with the minimum requirements that doctors and healthcare professionals consider when a person is piped. Also, it helps in the seek to comply the regulations that the federal government agency elaborates with the purpose of reviewing the existing proposals for open source mechanical ventilators. It also contains the technical requirements that are need to be covered by the designers. These regulations cover the feasibility for replicating the ventilators proposed, based on certain factors that will be described in this paper. Once the ventilators have been tested, its improvement is carried out from the mechanical part, considering the electrical element to be used, in order to obtain a transitional mechanical ventilator that could be easily replicated with national suppliers.

Higher Education Institutions implement study programs with the aim of training competitive professionals that adjust to the needs of the labor market, but seeking as far as possible to guarantee a successful graduation. It is necessary to identify the characteristics of graduation of a program education that allows institutional managers to propose strategies that improve graduation levels in them and thereby improve the quality conditions offered to students enrolled in these study programs. In order to carry out this process, the failure rate, the conditions of the school environment, the results of the bachelor's exit exams and the accreditations of the educational program are analyzed from the beginning. This is done in order to identify key factors in the dropout process and where to work for the improvement of the Electronic Engineering educational program of the Faculty of Chemical Sciences and Engineering of the Autonomous University of Baja California.