| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 5.03 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
O ser humano possui a incrível capacidade de contagiar-se com emoções de outros indivíduos.
Este projeto tem o intuito de auxiliar esta transferência de felicidade e animação. Com a utilização de um acelerômetro preso a mão do utilizador, realiza o mapeamento dos movimentos e os transmite via bluetooth para um microcomputador, onde é classificado os movimentos e os transforma em mensagens DMX transmitidas via USB para um projetor holográfico, gerando assim padrões luminosos, sendo possível dessa maneira expressar a agitação do utilizador. Para isto utilizou-se a técnica de Aprendizagem de maquina para a classificação dos movimentos, mais precisamente o algoritmo de Regressão Logistica, que ainda foi comparado com outro algoritmo, o Support Vector Machine. Por fim os resultados foram satisfatórios alcançando uma precisão de aproximadamente 80%.
The human being possesses the incredible capacity to be infected with the emotions of other individuals. This project is intended to help this transfer of happiness and animation. With the use of an accelerometer attached to the user's hand, it maps the movements and transmits them via bluetooth to a microcomputer, where the movements are classified and transformed into DMX messages transmitted via USB to a holographic projector, thus generating luminous patterns, being able in this way, to express the agitation of the user. For this purpose, the Machine Learning technique was used to classify movements, more precisely the Logistic Regression algorithm, which was still compared to another algorithm, the Support Vector Machine. Finally the results were satisfactory reaching an accuracy of approximately 80%.
The human being possesses the incredible capacity to be infected with the emotions of other individuals. This project is intended to help this transfer of happiness and animation. With the use of an accelerometer attached to the user's hand, it maps the movements and transmits them via bluetooth to a microcomputer, where the movements are classified and transformed into DMX messages transmitted via USB to a holographic projector, thus generating luminous patterns, being able in this way, to express the agitation of the user. For this purpose, the Machine Learning technique was used to classify movements, more precisely the Logistic Regression algorithm, which was still compared to another algorithm, the Support Vector Machine. Finally the results were satisfactory reaching an accuracy of approximately 80%.
Description
Mestrado de dupla diplomação com a UTFPR - Universidade Tecnológica Federal do Paraná
Keywords
DMX OLA Raspberrypi Acelerômetro Machine learning Bluetooth.