Face Features Recognition

Subject to your needs, you can enable automatic recognition of such face features as gender, age, emotions, glasses, beard, and face mask. This functionality can be activated on both GPU- and CPU-accelerated video face detectors.

In this section:

Enable Face Features Recognition

Important

This step will enable face features recognition via HTTP API.

To enable automatic recognition of face features, open the /etc/findface-extraction-api configuration file and enable relevant recognition models: gender, age, emotions, glasses, beard, and face mask. Be sure to choose the right acceleration type for each model, matching the acceleration type of findface-extraction-api: CPU or GPU. Be aware that findface-extraction-api on CPU can work only with CPU-models, while findface-extraction-api on GPU supports both CPU- and GPU-models.

sudo vi /etc/findface-extraction-api.ini

models:
  age: faceattr/age.v1.cpu.fnk
  emotions: faceattr/emotions.v1.cpu.fnk
  face: face/ifruit_320.cpu.fnk
  gender: faceattr/gender.v2.cpu.fnk
  beard: faceattr/beard.v0.cpu.fnk
  glasses3: faceattr/glasses3.v0.cpu.fnk
  medmask3: faceattr/medmask3.v2.cpu.fnk

The following models are available:

Note

You can find face features recognition models at /usr/share/findface-data/models/faceattr/.

ls /usr/share/findface-data/models/faceattr/
age.v1.cpu.fnk  age.v1.gpu.fnk  beard.v0.cpu.fnk  beard.v0.gpu.fnk  emotions.v1.cpu.fnk  emotions.v1.gpu.fnk  gender.v2.cpu.fnk  gender.v2.gpu.fnk  glasses3.v0.cpu.fnk  glasses3.v0.gpu.fnk medmask3.v2.cpu.fnk medmask3.v2.gpu.fnk liveness.v3.gpu.fnk quality.v1.cpu.fnk quality.v1.gpu.fnk

Face feature	Acceleration	Configuration file parameter
face (biometry)	CPU	face: face/ifruit_320.cpu.fnk face: face/ifruit_160.cpu.fnk
	GPU	face: face/ifruit_320.gpu.fnk face: face/ifruit_160.gpu.fnk
age	CPU	age: faceattr/age.v1.cpu.fnk
	GPU	age: faceattr/age.v1.gpu.fnk
gender	CPU	gender: faceattr/gender.v2.cpu.fnk
	GPU	gender: faceattr/gender.v2.gpu.fnk
emotions	CPU	emotions: faceattr/emotions.v1.cpu.fnk
	GPU	emotions: faceattr/emotions.v1.gpu.fnk
glasses	CPU	glasses3: faceattr/glasses3.v0.cpu.fnk
	GPU	glasses3: faceattr/glasses3.v0.gpu.fnk
beard	CPU	beard: faceattr/beard.v0.cpu.fnk
	GPU	beard: faceattr/beard.v0.gpu.fnk
face mask	CPU	medmask3: faceattr/medmask3.v2.cpu.fnk
	GPU	medmask3: faceattr/medmask3.v2.gpu.fnk

Tip

To disable a recognition model, simply pass an empty value to a relevant parameter. Do not remove the parameter itself as in this case the system will be searching for the default model.

models:
  gender: ""
  age: ""
  emotions: ""

Restart findface-extraction-api.

sudo systemctl restart findface-extraction-api

Once the models are enabled, be sure to configure the web interface to display the recognition results.

Display Features Recognition Results in Events

To display the face features recognition results in the event list, add the following line into the FFSECURITY section: 'EVENTS_FEATURES': ['gender', 'age', 'emotions', 'beard', 'glasses', 'medmask'], subject to the list of enabled models.

Warning

This line must be placed between SF_API_ADDRESS and LIVENESS_THRESHOLD as shown in the example.

sudo vi /etc/findface-security/config.py

...
FFSECURITY = {
...
'SF_API_ADDRESS': 'http://127.0.0.1:18411',
'EVENTS_FEATURES': ['gender', 'age', 'emotions', 'beard', 'glasses', 'medmask'],
'LIVENESS_THRESHOLD': 0.75,
'BEARD_THRESHOLD': 0.7,
 }

Restart findface-security.

sudo systemctl restart findface-security

Face Features in Events

Once the face features recognition configured, you will see the recognition result for each found face in the following format:

Face feature	Result format	Example
Age	Feature: age: number of years	age: 33
Gender	Result: male/female (feature: gender): algorithm confidence in result	female (gender): 0.95
Emotions	Result: angry/disgust/fear/happy/sad/surprise (feature: emotions): algorithm confidence in result	happy (emotions): 0.99
Glasses	Result: eye/sun/none (feature: glasses): algorithm confidence in result	none (glasses): 0.87
Beard	Result: beard/none (feature: beard): algorithm confidence in result	none (beard): 0.91
Face mask	Result: none/correct (feature: face mask): algorithm confidence in result	none (face mask): 0.93

Filter events by face features when needed.