1. Mining and preprocessing of training data

2. Use the mainstream deep learning framework to realize the recognition and pixel segmentation of specific objects in the image

3. Combining deep learning and traditional visual tracking algorithms to achieve tracking and trajectory analysis of specific targets in images

1. More than two years of work experience, major in computer science or related, bachelor degree or above, master degree is preferred;

2. Familiar with mainstream deep learning frameworks, such as Caffe, TensorFlow, MXNet, and have strong Python/C++ programming and development capabilities