Interpretable Latent Spaces for Learning from Demonstration
This addresses the challenge of human-robot interaction by improving interpretability in robot learning from human demonstration, though it is incremental as it builds on existing latent representation methods.
The paper tackles the problem of making latent representations in learning from demonstration interpretable to humans by aligning them with user-defined symbols, achieving alignment in both simulated and real-world object data experiments.
Effective human-robot interaction, such as in robot learning from human demonstration, requires the learning agent to be able to ground abstract concepts (such as those contained within instructions) in a corresponding high-dimensional sensory input stream from the world. Models such as deep neural networks, with high capacity through their large parameter spaces, can be used to compress the high-dimensional sensory data to lower dimensional representations. These low-dimensional representations facilitate symbol grounding, but may not guarantee that the representation would be human-interpretable. We propose a method which utilises the grouping of user-defined symbols and their corresponding sensory observations in order to align the learnt compressed latent representation with the semantic notions contained in the abstract labels. We demonstrate this through experiments with both simulated and real-world object data, showing that such alignment can be achieved in a process of physical symbol grounding.