ImpulseRunner

class edge_impulse_linux.runner.ImpulseRunner(
		model_path: str
)

Parameters

• model_path: str

Subclasses

• edge_impulse_linux.audio.AudioImpulseRunner

• edge_impulse_linux.image.ImageImpulseRunner

Methods

classify

edge_impulse_linux.runner.classify(
		self,
		data
)

Parameters

• self

• data

hello

edge_impulse_linux.runner.hello(
		self
)

Parameters

• self

init

edge_impulse_linux.runner.init(
		self,
		debug=False
)

Parameters

• self

• debug=False

send_msg

edge_impulse_linux.runner.send_msg(
		self,
		msg
)

Parameters

• self

• msg

stop

edge_impulse_linux.runner.stop(
		self
)

Parameters

• self

now

edge_impulse_linux.runner.now(
		
)

Submodules

• edge_impulse_linux.audio

• edge_impulse_linux.image

• edge_impulse_linux.runner

Package Info

• Name: edge_impulse_linux

• Summary: Python runner for real-time ML classification

• Version: 1.0.10

• Requires: numpy, psutil

• License: None

AudioImpulseRunner

Parameters

• model_path: str

Bases

• edge_impulse_linux.runner.ImpulseRunner

Methods

classifier

Parameters

• self

• device_id=None

classify

Parameters

• self

• data

init

Parameters

• self

• debug=False

Microphone

Parameters

• rate

• chunk_size

• device_id=None

• channels=1

Methods

checkDeviceModelCompatibility

Parameters

• self

• device_id

fill_buffer

Parameters

• self

• in_data

• frame_count

• time_info

• status_flags

generator

Parameters

• self

listAvailableDevices

Parameters

• self

now

ImageImpulseRunner

Parameters

• model_path: str

Bases

• edge_impulse_linux.runner.ImpulseRunner

Methods

classifier

Parameters

• self

• videoDeviceId=0

classify

Parameters

• self

• data

get_features_from_image

Parameters

• self

• img

• crop_direction_x='center'

• crop_direction_y='center'

get_features_from_image_auto_studio_setings

Parameters

• self

• img

get_frames

Parameters

• self

• videoDeviceId=0

init

Parameters

• self

• debug=False

get_features_from_image_with_studio_mode

Extract features from an image using different resizing modes suitable for Edge Impulse Studio.

Parameters

• img

• mode

• output_width

• output_height

• is_grayscale

resize_with_letterbox

Resize an image while maintaining aspect ratio using letterboxing.

Parameters

• image

• target_width

• target_height