2018-10-23-hristov18a.md

title	abstract	keywords	layout	series	id	month	tex_title	firstpage	lastpage	page	order	cycles	bibtex_author	author	date	address	publisher	container-title	volume	genre	issued	pdf	extras
Interpretable Latent Spaces for Learning from Demonstration	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.	disentanglement learning, model interpretability, symbol grounding	inproceedings	Proceedings of Machine Learning Research	hristov18a	0	Interpretable Latent Spaces for Learning from Demonstration	957	968	957-968	957	false	Hristov, Yordan and Lascarides, Alex and Ramamoorthy, Subramanian	given family Yordan Hristov given family Alex Lascarides given family Subramanian Ramamoorthy	2018-10-23		PMLR	Proceedings of The 2nd Conference on Robot Learning	87	inproceedings	date-parts 2018 10 23	http://proceedings.mlr.press/v87/hristov18a/hristov18a.pdf	label link Source code https://github.com/yordanh/interp_latent_spaces