Open for customers to label objects and movements in video at will
Labeling software is a crucial part of artificial intelligence. The tool decides the diversity and accuracy of object detection, and YUAN owns a set of labeling software developed entirely on our own for customers to train models they need. Customers can organize their labeling team and collect video data on their own. Users can define objects needed to be detected and label those data no matter if those are from confidential sources or for customized projects.
More importantly, the labeling software integrates the function to train a deep learning model. GPU takes in video data labeled at the software and generates data file that can be copied directly to QDEEP SDK after computing and learning. Our customers can build an AI deep learning system with our labeling tool.
QDEEP SDK provides various functions software developers frequently require, and it meanwhile supports the following hardware and deep learning models.
Support Hardware: nVIDIA CUDA , Intel Movidius
Deep Learning Architecture: TensorFlow , CAFFE