Develop a machine learning model to predict places with high risk of sexual harassment incidents against women on India. The project was made in contribution with Safecity’s website.

The abstract for my master thesis:

Organ segmentation of the human body in medical imaging is a widely used process not only in medicine for detection and diagnosis of diseases, as well as in education helping students in learning the human anatomy. Despite its enormous importance, this process is very time-consuming and costly as it requires experts in the field, time and adequate tools to do so. With the advances of the artificial intelligence, the goal of this work is to evaluate the automatic bone segmentation in computed tomography (CT) images using deep learning networks on the full body databases provided by the Visible Human Project (VHP). The segmentation is done by each model classifying each pixel from the tomography image into one of the following 19 classes (18 bones + background) being: cranium, jaw, clavicle, scapula, humerus, radius, ulna, hands, ribs, sternum, vertebras, sacrum, pelvis, femur, patella , tibia, fibula and feet.

The proposed method consists of using the male body tomographic base containing about 1865 images of size 512 x 512 to train various deep learning network architectures in supervised mode including data augmentation techniques and external databases aiming to test its generalization for the bone segmentation on the female body base containing close to 1730 images. The quantitative evaluation of the various network topologies at the end of the process will be made using the Dice coefficient as metric.

Project dedicate to collect, store and process various sensor feebacks from the automation field of the production lines on the industry and compile it in usefull information for use from field operators to top managers on decision making.

Tasks I developed:

* Capture in real time data from various types of sensors (PLC, C++)

* Process and storethe collect data in the appropriate database (ORACLE, MySQL)

* Compile the collected data into usefull information for various staff members ranging from machine operators to top managment helping them taking important actions to increase productivity (PHP, HTML, Javascript, CSS)

Project responsible for storing and compiling information from the assets provided by the technical maintenance helping taking early decisions to avoid machine breaks.

Tasks I developed:

* Create database system for storing various types of reports (MySQL)

* Collect data using OPC technology from field sensors and storing them in real time on a database (PLC, OPC, SCADA)

* Provide in a quick and meaninfull way insights of the current states of the assets on the production lines (PHP, HTML, JavaScript, CSS)

Project responsable for storing information collected by the laboratory staff when analyzing steel samples from the production lines on the siderurgyproviding important feedback for the operational staff to make the proper machinery adjustments.

Tasks I developed:

* Connect different databases to sycronize data from various sources(ORACLE, C++, PLC, SCADA, OPC)

* Create interface for laboratory staff to insert important information from the samples analysed. (Flex, PHP, HTML)

* Create interface for operational staff to check in real time the information inserted by the laboratory staff to take actions for correcting process deviations.(Flex, PHP, HTML)

Updated obsolete SCADA system from automation for various production lines.

Tasks I developed:

* Updated windows operation system from WinNT to Windows 2003.

* Updated Siemens's SCADA WinCC.

Developing a project based on computer vision for a autonomous car on the city of Curitiba.

Tasks I developed:

* Lane Detection

* Traffic Sign Detection

Development of automation projects focused on reducing problems of steel breaking, machine breakdown and increased productivity of the Cold Rolling Mill and Coil Coating Lines.

Tasks I developed:

* Software (WinCC, Step7, TDC) and hardware (PLC, Sensors, Drivers)improvements on the Level 1 automation system provided by Siemens.

* Software (C++) and hardware (Windows Server) improvements on the Leve 2 automation system that is responsable for stablishing the communication from MES (Manufacturing Execution System)with Level 1.