We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations

DBLP - CS Bibliography


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Machine Learning

Title: Translating Natural Language Instructions for Behavioral Robot Navigation with a Multi-Head Attention Mechanism

Abstract: We propose a multi-head attention mechanism as a blending layer in a neural network model that translates natural language to a high level behavioral language for indoor robot navigation. We follow the framework established by (Zang et al., 2018a) that proposes the use of a navigation graph as a knowledge base for the task. Our results show significant performance gains when translating instructions on previously unseen environments, therefore, improving the generalization capabilities of the model.
Comments: Accepted at ACL 2020 WiNLP workshop
Subjects: Machine Learning (cs.LG); Computation and Language (cs.CL); Robotics (cs.RO); Machine Learning (stat.ML)
Cite as: arXiv:2006.00697 [cs.LG]
  (or arXiv:2006.00697v3 [cs.LG] for this version)

Submission history

From: Patricio Cerda-Mardini [view email]
[v1] Mon, 1 Jun 2020 03:49:43 GMT (120kb,D)
[v2] Tue, 2 Jun 2020 19:49:53 GMT (119kb,D)
[v3] Sun, 7 Jun 2020 23:00:47 GMT (119kb,D)

Link back to: arXiv, form interface, contact.