Sensors Cortical Areas
Welcome to the sensors cortical areas type tutorial! Other than vision that was covered in the previous tutorial, we support many other types. Some of the sensors in FEAGI include infrared, inverse infrared, proximity, accelerometer, gyroscope, battery gauge, central vision, and peripheral visions.
Title and Sensors Chart
Welcome to the sensors cortical areas type tutorial! Other than vision that was covered in the previous tutorial, we support many other types. Some of the sensors in FEAGI include infrared, inverse infrared, proximity, accelerometer, gyroscope, battery gauge, central vision, and peripheral visions.
Infrared and Reverse Infrared
An infrared sensor is a device that emits and/or detects infrared radiation to sense its surroundings. Infrared radiation is a form of light that exists just beyond the visible to the human eye spectrum. In FEAGI, an infrared input cortical area helps activate neurons when an infrared sensor is activated. According to how many infrared sensors your robot embodiment has, you can extend the infrared cortical area across the x axis. In this example, this robot has 3 infrared sensors, so we extended the infrared cortical area to be 3 blocks wide. Here is the physical representation of that robot. When you wave an object that blocks the infrared signals, you can see the infrared cortical areas deactivate in FEAGI. Inverse infrared cortical areas have the same dimensions, but they serve the opposite purpose of regular infrared sensors. Instead, inverse infrared cortical areas activate neurons when there is a lack of activation in the infrared sensor. You can see in this example that when something covers the infrared sensor, the cortical area is activated. Inverse infrared cortical areas are also extendable according to how many sensors your robot embodiment has.
Proximity
The proximity sensor has default dimensions of 1 x 1 x 10, with each block representing a number on a incremental scale of 0-9. The proximity sensor is responsible for encoding distance measurement signals. 0 indicates that the object is very close, while 9 indicates that the object is far away. In this physical robot embodiment, as we move the object farther and closer away, you can see the cortical area activation move back and forth along the cortical area.
Acceleometer
An accelerometer is responsible for detecting the rate of change for x, y, z movements. The accelerometer cortical area has default dimensions of 3 x 1 x 21. The length of the y axis acts as a scale from -100% to positive 100%. A number on the furthest end of the cortical area indicates the highest rate of acceleration change, while 0 indicates the speed is constant. Values in the middle correspond their number accordingly, like positive or negative 10% indicating the lowest rate of change. Each column in the accelerometer acts as x, y, and z, axis indicators. Together, they work together to encode accelerometer signals in their respective x, y, z coordinates. For the first column, the negative and positive values capture the magnitude of the robot acceleration across left and right movements. -10% and + 10% indicate slower movements of left and right, respectively 100% and -100% indicate a greater rate of change.
Gyroscope
Similar to accelerometer, the gyroscope cortical area has default dimensions of 3 x1 x 21. The length of the y axis acts as a scale from -100% to positive 100%. A number on the furthest end of the cortical area indicates the maximum angular rotation, while 0 indicates the robot is level. Values in the middle correspond to their number accordingly, like positive or negative 10% indicating minimal rotation. Each column in the gyroscope acts as roll, pitch, and yaw indicators. In this drone embodiment example, you can see that negative and positive values in the 1st column captures the rotation of robot acceleration for clockwise and counterclockwise roll movements. -10% and +10% indicate slower movements, while 100% and -100% indicate faster movements. For the second column, the negative and positive values capture the rotation of the robot for forward and backward pitch movements. For the third column, the negative and positive values capture the rotation of the robot for left and right yaw movements. Adding accelerometer and gyroscope cortical areas to your experiment allows you to monitor the acceleration and angular orientation of different movements.
Battery Gauge
The battery gauge represents the available battery level. The default dimensions of 1 x 1 x 10 indicates the power on a scale of 0-9. A value of 0 means there is not power, but a value of 9 means there is full power available.
Central Vision
The central vision cortical area interprets visual signals from a centralized view. By default, it will be in grayscale with dimensions of 64 x 64 x 1. If you would like your vision include colors, you can extend the z dimensions by 3, representing a human eye's red, green, and blue color sensing cones.
Peripheral Vision
The peripheral vision sensors interpret visual signals from specific views. Respectively, these include top right, top left, top middle, middle left, middle right, bottom left, bottom right, and bottom middle. Together, the peripheral and central visions encode visual stimuli. In this example of pong in FEAGI, you can see the ball moving in the central vision. When the ball enters the respective section of the peripheral vision's section, you can see the cortical areas activate.
