Compositional Obverter Communication Learning From Raw Visual Input

# Compositional Obverter Communication Learning From Raw Visual Input

**arXiv ID:** 1804.02341
**Authors:** Edward Choi, Angeliki Lazaridou, Nando de Freitas
**Published:** 2018-04-06T16:12:51Z
**Abstract:**
One of the distinguishing aspects of human language is its compositionality, which allows us to describe complex environments with limited vocabulary. Previously, it has been shown that neural network agents can learn to communicate in a highly structured, possibly compositional language based on disentangled input (e.g. hand- engineered features). Humans, however, do not learn to communicate based on well-summarized features. In this work, we train neural agents to simultaneously develop visual perception from raw image pixels, and learn to communicate with a sequence of discrete symbols. The agents play an image description game where the image contains factors such as colors and shapes. We train the agents using the obverter technique where an agent introspects to generate messages that maximize its own understanding. Through qualitative analysis, visualization and a zero-shot test, we show that the agents can develop, out of raw image pixels, a language with compositional properties, given a proper pressure from the environment.

## Skill Description

This skill is generated from the arXiv paper: Compositional Obverter Communication Learning From Raw Visual Input (1804.02341).

## How to Use

[To be filled in by the user or by future automation]

## References

- [arXiv:1804.02341](http://arxiv.org/abs/1804.02341v1)