Our paper "Hierarchical Interest-Driven Associative Goal Babbling for Efficient Bootstrapping of Sensorimotor Skills" by Rania Rayyes, Heiko Donat and Jochen Steil, has been accepted at IEEE International Conference on Robotics and Automation (ICRA) 2020. The paper presents efficient online data-driven learning scheme for developmental robotics which is demonstrated on 7 DoF Baxter. The robot is intrinsically motivated to explore its workspace. The paper proposed the first intrinsic motivation method which combines knowledge-based and competence-based intrinsic signals and demonstrates its efficiency where only 2 hours of direct training on real robot was required to learn the task. It also proposed the first Associative neural network to be constructed totally from scratch with high stability demonstrated in the real robot experiment. Here is a short video about the work