Face Detection
Detection of faces in still images or videos
The face detection mining service allows to detect one or more faces to be found in images and videos.
Input: Image file or video file. You can specify which frames are to be processed for a video.
Output: A set of detected faces will be returned for the image or each processed frame. For each detected face an axially parallel bounding box and a vectorized representation of the face (embedding) are returned. In addition, an automatically generated ID is assigned to each detected face to allow the unambiguous identification of all detected faces in one media file.
The service utilizes the Dlib toolkit for machine learning. Image processing in Dlib is based on HoG (Histograms of Oriented Gradients for Human Detection). Dlib implements a front face detector which uses a pretrained network for the recognition of human faces. The network has been trained on 3 million faces derived from a number of datasets: The face scrub dataset, the VGG dataset, and a large number of images Davis King, the creator of Dlib, scraped from the internet.
The mining service is also able to recognize faces from a previously trained gallery of people's faces. If a detected face matches a previously trained one, the mining result contains a label (e.g. Name or URI) and a score representing the confidence of the assignment.