2023-12-02-krupnik23a.md

title	section	openreview	abstract	layout	series	publisher	issn	id	month	tex_title	firstpage	lastpage	page	order	cycles	bibtex_author	author	date	address	container-title	volume	genre	issued	pdf	extras
Fine-Tuning Generative Models as an Inference Method for Robotic Tasks	Poster	yGkqN4hqrJ	Adaptable models could greatly benefit robotic agents operating in the real world, allowing them to deal with novel and varying conditions. While approaches such as Bayesian inference are well-studied frameworks for adapting models to evidence, we build on recent advances in deep generative models which have greatly affected many areas of robotics. Harnessing modern GPU acceleration, we investigate how to quickly adapt the sample generation of neural network models to observations in robotic tasks. We propose a simple and general method that is applicable to various deep generative models and robotic environments. The key idea is to quickly fine-tune the model by fitting it to generated samples matching the observed evidence, using the cross-entropy method. We show that our method can be applied to both autoregressive models and variational autoencoders, and demonstrate its usability in object shape inference from grasping, inverse kinematics calculation, and point cloud completion.	inproceedings	Proceedings of Machine Learning Research	PMLR	2640-3498	krupnik23a	0	Fine-Tuning Generative Models as an Inference Method for Robotic Tasks	866	886	866-886	866	false	Krupnik, Orr and Shafer, Elisei and Jurgenson, Tom and Tamar, Aviv	given family Orr Krupnik given family Elisei Shafer given family Tom Jurgenson given family Aviv Tamar	2023-12-02		Proceedings of The 7th Conference on Robot Learning	229	inproceedings	date-parts 2023 12 2	https://proceedings.mlr.press/v229/krupnik23a/krupnik23a.pdf