Self-Supervised Interpretable End-to-End Learning via Latent Functional Modularity

Hyunki Seong, Hyunchul Shim
Proceedings of the 41st International Conference on Machine Learning, PMLR 235:44212-44223, 2024.

Abstract

We introduce MoNet, a novel functionally modular network for self-supervised and interpretable end-to-end learning. By leveraging its functional modularity with a latent-guided contrastive loss function, MoNet efficiently learns task-specific decision-making processes in latent space without requiring task-level supervision. Moreover, our method incorporates an online, post-hoc explainability approach that enhances the interpretability of end-to-end inferences without compromising sensorimotor control performance. In real-world indoor environments, MoNet demonstrates effective visual autonomous navigation, outperforming baseline models by 7% to 28% in task specificity analysis. We further explore the interpretability of our network through post-hoc analysis of perceptual saliency maps and latent decision vectors. This provides valuable insights into the incorporation of explainable artificial intelligence into robotic learning, encompassing both perceptual and behavioral perspectives. Supplementary materials are available at https://sites.google.com/view/monet-lgc.

Cite this Paper

BibTeX
@InProceedings{pmlr-v235-seong24a, title = {Self-Supervised Interpretable End-to-End Learning via Latent Functional Modularity}, author = {Seong, Hyunki and Shim, Hyunchul}, booktitle = {Proceedings of the 41st International Conference on Machine Learning}, pages = {44212--44223}, year = {2024}, editor = {Salakhutdinov, Ruslan and Kolter, Zico and Heller, Katherine and Weller, Adrian and Oliver, Nuria and Scarlett, Jonathan and Berkenkamp, Felix}, volume = {235}, series = {Proceedings of Machine Learning Research}, month = {21--27 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v235/main/assets/seong24a/seong24a.pdf}, url = {https://proceedings.mlr.press/v235/seong24a.html}, abstract = {We introduce MoNet, a novel functionally modular network for self-supervised and interpretable end-to-end learning. By leveraging its functional modularity with a latent-guided contrastive loss function, MoNet efficiently learns task-specific decision-making processes in latent space without requiring task-level supervision. Moreover, our method incorporates an online, post-hoc explainability approach that enhances the interpretability of end-to-end inferences without compromising sensorimotor control performance. In real-world indoor environments, MoNet demonstrates effective visual autonomous navigation, outperforming baseline models by 7% to 28% in task specificity analysis. We further explore the interpretability of our network through post-hoc analysis of perceptual saliency maps and latent decision vectors. This provides valuable insights into the incorporation of explainable artificial intelligence into robotic learning, encompassing both perceptual and behavioral perspectives. Supplementary materials are available at https://sites.google.com/view/monet-lgc.} }
Endnote
%0 Conference Paper %T Self-Supervised Interpretable End-to-End Learning via Latent Functional Modularity %A Hyunki Seong %A Hyunchul Shim %B Proceedings of the 41st International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2024 %E Ruslan Salakhutdinov %E Zico Kolter %E Katherine Heller %E Adrian Weller %E Nuria Oliver %E Jonathan Scarlett %E Felix Berkenkamp %F pmlr-v235-seong24a %I PMLR %P 44212--44223 %U https://proceedings.mlr.press/v235/seong24a.html %V 235 %X We introduce MoNet, a novel functionally modular network for self-supervised and interpretable end-to-end learning. By leveraging its functional modularity with a latent-guided contrastive loss function, MoNet efficiently learns task-specific decision-making processes in latent space without requiring task-level supervision. Moreover, our method incorporates an online, post-hoc explainability approach that enhances the interpretability of end-to-end inferences without compromising sensorimotor control performance. In real-world indoor environments, MoNet demonstrates effective visual autonomous navigation, outperforming baseline models by 7% to 28% in task specificity analysis. We further explore the interpretability of our network through post-hoc analysis of perceptual saliency maps and latent decision vectors. This provides valuable insights into the incorporation of explainable artificial intelligence into robotic learning, encompassing both perceptual and behavioral perspectives. Supplementary materials are available at https://sites.google.com/view/monet-lgc.
APA
Seong, H. & Shim, H.. (2024). Self-Supervised Interpretable End-to-End Learning via Latent Functional Modularity. Proceedings of the 41st International Conference on Machine Learning, in Proceedings of Machine Learning Research 235:44212-44223 Available from https://proceedings.mlr.press/v235/seong24a.html.

Related Material