Sign inSign up
Jointly Learning to Parse and Perceive: Connecting Natural Language to the Physical World thumbnail
Pause
Mute
Subtitles
Playback speed
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Full screen

Jointly Learning to Parse and Perceive: Connecting Natural Language to the Physical World

Published on Oct 02, 20132312 Views

Jayant Krishnamurthy

This paper introduces Logical Semantics with Perception (LSP), a model for grounded language acquisition that learns to map natural language statements to their referents in a physical environment. Fo

ACL 2013 - Sofia

Related categories

Computational Linguistics

Chapter list

Jointly Learning to Parse and Perceive: Connecting Natural Language to the Physical World00:00
Grounded Language Understanding00:19
Grounding Natural Language Descriptions01:27
Outline - 103:07
Logical Semantics with Perception (LSP) - 103:29
Logical Semantics with Perception (LSP) - 203:40
Logical Semantics with Perception (LSP) - 304:17
Logical Semantics with Perception (LSP) - 404:46
Logical Semantics with Perception (LSP) - 505:02
LSP in Math05:27
Perception06:21
Outline - 207:47
Weakly-Supervised Training07:55
Parameter Estimation - 108:24
Parameter Estimation - 211:03
Outline - 311:47
Data sets12:04
Evaluation Metric14:12
Results (Scene)15:12
Results (GeoQA)19:18
Contributions19:50