igibson.sensors package

Submodules

igibson.sensors.bump_sensor module

class igibson.sensors.bump_sensor.BumpSensor(env)

Bases: igibson.sensors.sensor_base.BaseSensor

Bump sensor

get_obs(env)

Get Bump sensor reading

Returns: Bump sensor reading

igibson.sensors.dropout_sensor_noise module

class igibson.sensors.dropout_sensor_noise.DropoutSensorNoise(env)

Bases: igibson.sensors.sensor_noise_base.BaseSensorNoise

Naive dropout sensor noise model

add_noise(obs)

Add naive sensor dropout to perceptual sensor, such as RGBD and LiDAR scan

Parameters

sensor_reading – raw sensor reading, range must be between [0.0, 1.0]
noise_rate – how much noise to inject, 0.05 means 5% of the data will be replaced with noise_value
noise_value – noise_value to overwrite raw sensor reading

Returns

sensor reading corrupted with noise

set_noise_rate(noise_rate)

Set noise rate

Parameters: noise_rate – noise rate

set_noise_value(noise_value)

Set noise value

Parameters: noise_value – noise value

igibson.sensors.scan_sensor module

igibson.sensors.sensor_base module

class igibson.sensors.sensor_base.BaseSensor(env)

Bases: object

Base Sensor class. Sensor-specific get_obs method is implemented in subclasses

abstract get_obs(env)

Get sensor reading

Param: environment instance
Returns: observation (numpy array or a dict that maps str to numpy array)

igibson.sensors.sensor_noise_base module

class igibson.sensors.sensor_noise_base.BaseSensorNoise(env)

Bases: object

Base SensorNoise class. Sensor noise-specific add_noise method is implemented in subclasses

abstract add_noise(obs)

Add sensor noise to sensor reading

Parameters: obs – raw observation
Returns: observation with noise

igibson.sensors.velodyne_sensor module

class igibson.sensors.velodyne_sensor.VelodyneSensor(env)