My team has developed neural network programs for 10 years. We has experienced adaboost and SVM,an now we are using deep learning. We have made many artificial intelligence programs like face recognition program, red eye recognition, fingerprint detection, abnormal motion detection programs.
We are using deep neural network (Deep Learning) technology and are using the Tensorflow framework (Tensorflow 12.0) and the Caffe framework for its implementation.
We now use deep neural networks to produce gender, age and emotion recognition products with human faces, and gender and age products are considered to have attained a satisfactory level.
The image processing technology currently retained by the development team includes the recognition areas of face recognition, car number recognition, character recognition, fingerprint recognition, counting areas of vehicle and personnel counting, vehicle and population density testing, and products for object detection and tracking. In addition to the above-mentioned deep neural networks, we use various neural network techniques, SVM, Adaboost, and various learners for image training to perform recognition, counting and detection.
Also, products of papers presented in various fields such as medical field, including abnormal motion detection, respiratory number detection using human body micro-vibration detection, and face-based heart beat detection, are also implemented and applicable to variety of image processing papers.