Fovea reports the bounding box for each detected face in a four field format, described below. These bounding boxes can be easily plotted over the original image, with ImageMagick, for instance.
- Left-X
- Top-Y
- Width
- Height
Google: [user@host]$ fovea --google --faces 7.png
Microsoft: [user@host]$ fovea --microsoft --faces 7.png
OpenCV: [user@host]$ fovea --opencv --faces 7.png
Amazon: [user@host]$ fovea --amazon --faces 7.png
Clarifai: [user@host]$ fovea --clarifai --faces 7.png
Watson: [user@host]$ fovea --watson --faces 7.png