Day 86 – Deep Learning Image Face Detection In Less Than A Min Using RetinaFace

Face detection, Facial recognition, and Facial expression recognition are popular nowadays after Artificial Intelligence.

iPhone was introduced on their phone, and many apps have these features.

I will show you how to detect face areas in the image in less than 25 lines of code.

I will use the RetinaFace DeepLearning library to detect all faces on the image.

What is RetinaFace?

RetinaFace is a deep learning based cutting-edge facial detector for Python coming with facial landmarks.

Source: RetinaFace

I am going to use Google Colab, and you can see the code here.

{‘score’: 0.9997192025184631, ‘facial_area’: [209, 83, 327, 250], ‘landmarks’: {‘right_eye’: [225.48073, 154.06593], ‘left_eye’: [277.90686, 148.1564], ‘nose’: [243.96909, 187.55666], ‘mouth_right’: [241.50496, 217.94543], ‘mouth_left’: [278.17957, 213.22757]}}
{‘score’: 0.9996291399002075, ‘facial_area’: [350, 95, 496, 300], ‘landmarks’: {‘right_eye’: [382.1355, 172.75745], ‘left_eye’: [445.83966, 169.52275], ‘nose’: [407.87024, 212.00275], ‘mouth_right’: [394.29565, 249.37596], ‘mouth_left’: [440.78555, 246.39479]}}
{‘score’: 0.9996178150177002, ‘facial_area’: [79, 100, 181, 242], ‘landmarks’: {‘right_eye’: [98.006485, 156.73376], ‘left_eye’: [147.02284, 155.76279], ‘nose’: [116.42045, 185.39218], ‘mouth_right’: [105.199776, 212.61584], ‘mouth_left’: [140.5728, 212.04578]}}
{‘score’: 0.9990742206573486, ‘facial_area’: [560, 104, 741, 342], ‘landmarks’: {‘right_eye’: [618.5633, 196.25557], ‘left_eye’: [700.45, 197.64638], ‘nose’: [663.01984, 244.12181], ‘mouth_right’: [627.3689, 287.5204], ‘mouth_left’: [687.887, 287.70456]}}
{‘score’: 0.9965476393699646, ‘facial_area’: [1, 81, 94, 209], ‘landmarks’: {‘right_eye’: [26.9971, 133.53793], ‘left_eye’: [67.458115, 129.10698], ‘nose’: [51.27739, 159.16124], ‘mouth_right’: [40.50979, 182.45554], ‘mouth_left’: [69.29862, 178.991]}}

Detect and extract individual faces from the image