... through direct compensation of relative errors. To achieve this, a novel calibration-driven transfer learning method is proposed for relative error prediction in dual-robot systems. Design/methodology/approach A novel local product of exponential (POE) model with minimal parameters is proposed...

.... The method uses end-to-end policy training to directly extract the optimal path from the visual inputs for the scene. Robotics Reinforcement learning Obstacle avoidance Transfer learning Human–robot interaction The overall distribution can be expressed as follows: (1) P ϕ ( λ...

... conversion and local feature extraction. Findings The experimental results show the effectiveness and practicability of the proposed method. Compared with the ordinary method, the proposed method gets much better metrics. Originality/value The originality is the transfer learning strategy that fine...