Continual Learning of Human-like Arm Postures

 

The 20th of August 2021, our method for continual learning of human-like arm postures has been published and released to the scientific community in the IEEE International Conference on Development and Learning (ICDL). This method takes inspiration from the learning experiential cycle that features humans in the process of getting familiar with the actions performed in similar situations in different temporal sessions. The proposed method incrementally train weight parameters and endows the robot with the capability of plan arm configurations faster session by session. 

I am happy to announce this publication and of having had the opportunity to present the result of this study in ICDL 2021. I believe that this paper shows significant preliminary results that can be of inspiration for future work in the field of human-like motion planning and learning.  

The abstract of the article is the following:

Inspired from established human motor control theories, our HUMP algorithm plans upper-limb collisions-free movements for anthropomorphic systems, which show kinematic human-like features [1]. Related cognitive issues can be further resolved when robots act as they are familiar with their workspace and can take initiative faster than in the early onsets of a task. Here, a continual learning technique is proposed to improve the performance of the HUMP under uncertainties of the items in a given scenario. Given the locality of the optimization-based HUMP algorithm, a meaningful initial guess, predicted from similar past motion experiences, can significantly reduce the computational cost and put the robot into action arguably faster than in the first attempts of planning with inexperienced initial guesses. This prediction is proposed to be incrementally refined by an optimal locally weighted regression method that operates on datasets of situational features that are regularly updated as new movements are planned by the robot in similar scenarios. The proposed cyclic experiential learner is tested on the selection of optimal human-like target postures in a reaching task with a large obstacle obstructing the straight-line path towards a given target. Results demonstrate the capability of extracting meaningful situational features in few sessions of online learning with a very limited size of the datasets. Comparisons with simple Euclidean locally weighted regression and random initializations showed the capability of planning target configurations of better quality with less computational cost. The proposed approach also exhibits to be robust against the interferences of new incoming samples depicting slightly changed situations of the same task.

DOI: 10.1109/ICDL49984.2021.9515565

Link to my oral presentation: https://youtu.be/9PULs6tp6hA