FaceServer

Set-up

# sudo ./nvinstall.sh
# sudo ./install.sh

Usage

# python3 main.py config/client.conf

Configurations

graphical
enabled	Enable GUI or not (not used).
listener
enabled	Enable listener or not.
listener_addr	Listening address of listener.
listener_port	Listening port of listener.
reporter
enabled	Enable reporter or not.
collector_addr	Address of remote collector
collector_port	Port of remote collector
collect_info_period	Period of collecting system-level and process-level information.
send_report_period	Period of sending report to collector

Handlers

ID		0
name		TxHandler
description		Transmit data through network
inputs	0	data to be transmit
parameters	addr	Target address
	port	Target port
	protocol	Via TCP / UDP

ID		1
name		RxHandler
description		Receive data from network
outputs	0	Received data.
parameters	addr	Listening address
	port	Listening port
	protocol	Via TCP / UDP

ID		2
name		CaptureHandler
description		Capture images from camera.
outputs	0	captured images.
parameters	source	Value 0 indicating the default building camera.
	width	Captured image width.
	height	Captured image height.
	fps	Frame rate of capturing images.

ID		3
name		RenderHandler
description		Render images to a window.
inputs	0	Images to be rendered.
parameters	fps	Frame rate of rendering images.

ID		4
name		ResizeHandler
description		To resize images.
inputs	0	Images to be resized.
outputs	0	Resized images.
parameters	width	Target width.
	height	Target height.

ID		5
name		EncodeHandler
description		Encode images to JPEG format
inputs	0	Images to be encoded.
outputs	0	Encoded images.
parameters	quality	Compression quality of JPEG.

ID		6
name		DecodeHandler
description		Decode compressed images
inputs	0	Images to be decoded.
outputs	0	Decoded images.

ID		7
name		YoloDetectObjectHandler
description		Detect object from images
inputs	0	Images
outputs	0	Locations of detected objects and their cropped images.
parameters	gpu	Enable GPU utilization or not.
	config	Model configurations.
	weights	Model weights.
	labels	Label of detected objects.
	targets	Interested objects.

ID		8
name		HaarDetectFaceHandler
description		Detect faces with Haar feature-based detector
inputs	0	images
outputs	0	Locations of detected faces and their cropped images.
parameters	scale_factor	Upsample scale of detection (the larger the faster, but less accurate).
	min_size	Minimum detect size.

ID		9
name		HogDetectFaceHandler
description		Detect faces with HoG-based detector
inputs	0	Images
outputs	0	Locations of detected faces and their cropped images.
parameters	upsample	Upsample times of detection (the larger the faster, but less accurate).

ID		10
name		MmodDetectFaceHandler
description		Detect faces with depp learning-based MMOD Model
inputs	0	Images
outputs	0	Locations of detected faces and their cropped images.
parameters	model	Model for detection
	upsample	Upsample times of detection (the larger the faster, but less accurate).
	tresh	Threshold of identifying faces.

ID		11
name		DlibAlignHandler
description		Align detected faces.
inputs	0	Cropped faces.
outputs	0	Aligned result
parameters	predictor	Model for alignment.
	indices	Features for alignment.
	dim	Resize size for alignment.

ID		12
name		MtcnnDetectFaceAlignHandler
description		Detect and align faces in an image
inputs	0	images
outputs	0	Location and crop of detected and aligned faces
parameters	model	Model for Detection.
	dim	Dimension of model (network size).
	margin	Padding of detected faces.

ID		13
name		DlibAlignRecognizeHandler
description		Align and recognize detected faces with Dlib model.
inputs	0	Cropped faces
outputs	0	Recongnized result (position & label)
parameters	predictor	Aligner.
	model	Model for recognition.
	classifier	Pretriained SVM classifier.
	dim	Dimenssion of model (network size).
	jitters	The larger the slower but more accurate.
	tresh	Threshold of recognition.

ID		14
name		OpenfaceRecognizeHandler
description		Recognize faces with OpenFace model.
inputs	0	Cropped faces
outputs	0	Recognized result (position & label)
parameters	model	Model for recognition.
	classifier	Pre-trained SVM classifier link here.
	dim	Dimension of model (network size).
	tresh	Threshold of recognition.

ID		15
name		FacenetRecognizeHandler
description		Recognize faces with FaceNet model.
inputs	0	Cropped faces
outputs	0	Recongnized result (position & label)
parameters	model	Model for recognition.
	classifier	Pretriained SVM classifier link here.
	dim	Dimension of model (network size).
	tresh	Threshold of recognition.

ID		16
name		DrawBoxesHandler
description		Draw boxes on image
inputs	0	Images.
	1	Locations of boxes.
outputs	0	Drawn images

ID		17
name		DrawLabeledBoxesHandler
description		Draw boxes and labels on image
inputs	0	Images.
	1	Locations of boxes with labels.
outputs	0	Drawn images