YOLOv3-416 is currently one of the best objection detection models and we customized it to perform face detection.Prior detection systems repurpose classifiers or localizers to perform detection. YOLOv3 uses a totally different approach. It applies a single neural network to the full image to achieve extraordinary results!
View DemoDespite significant recent advances in the field of face recognition, implementing face verification and recognition efficiently at scale presents serious challenges. FaceNet, tackles this by directly learning mappings from face images to a compact Euclidean space where distances directly correspond to a measure of face similarity.
View DemoAlthough, Hrithik Roshan is barely recognized in a few frames, he still is SEEN under that white mask posed as a sculpture from the Movie Dhoom 2 !
See how our model performs in recognizing the quick gun Amitabh Bachchan
Ocassionally, celebrities get lost in the crowd due to the huge number of fans that follow them. In such cases, having a model that could guide security personnel in locating and directing the VIP to a secure location would be really helpful.
Face recognition models can help identify suspects which can further help them crack cases faster. Devices such as body camera's provide as an ideal method for identification due to the clean angle of recording as seen in this video from Amar Akbar Anthony.
DISCLAIMER: NO HUMANS WERE HARMED DURING THE MAKING OF THIS VIDEO
Face recognition models can help identify perpetrators and victims involved in mob lynching. This can in the future help police solve mob lynching cases more efficiently and even prevent them!