My team is conducting various research related to robotics and machine vision. In one of our projects, we used DOPE pose estimation to determine the pose of power connectors, with which we achieved very good results. We thought about the industrial application of DOPE. However, according to the GitHub page, DOPE is licensed under the CC BY-NC-SA 4.0 license (GitHub - NVlabs/Deep_Object_Pose: Deep Object Pose Estimation (DOPE) – ROS inference (CoRL 2018)). (In contrast, PoseCNN is released under the MIT License.)
Our question would be: under what conditions can we use DOPE pose estimation in commercial applications?
Thank you for your answer!