Smart Vision Analysis

This project is related to neuroscience and explores brain computer interfacing. It aims to analyse the reception of images and colors to the brain to Improve computer vision techniques.

Methodology:

An experiment is designed to observe the brainwave changes for a subject, who is being subjected to changing visual stimuli. The observed brainwave changes are recorded,this data is then used to develop a machine learning model which will train itself to gain the ability of judging the visual stimulus the subject was subjected to by merely observing the subject’s brainwave data. This experiment will be further used to improve the efficiency of image perception and incorporate it with BCI .

Procedure:

The project consists of a few steps:
1.Using a Geodesic sensor net headset, we record data depicting the behaviour of brain nodes when shown different pictures and colors. Then we create an ML algoritm that tries the same sensory techniques used by the brain to detect the colors.
2. The ML algorithm is modified to now also analyse various objects in images.
3.This final ML algorithm that imitates the brain's behaviour while analysing objects is compared to other computer vision softwares.

Sensory Data:

The Y values here are the brain wave data that was recorded using the Geodesic sensor net, the observed values were converted to python arrays to plot them. The X values can be considered analogous to time.