Learning to Generate Compositional Color Descriptions

by   Will Monroe, et al.

The production of color language is essential for grounded language generation. Color descriptions have many challenging properties: they can be vague, compositionally complex, and denotationally rich. We present an effective approach to generating color descriptions using recurrent neural networks and a Fourier-transformed color representation. Our model outperforms previous work on a conditional language modeling task over a large corpus of naturalistic color descriptions. In addition, probing the model's output reveals that it can accurately produce not only basic color terms but also descriptors with non-convex denotations ("greenish"), bare modifiers ("bright", "dull"), and compositional phrases ("faded teal") not seen in training.


Learning Distributions of Meant Color

When a speaker says the name of a color, the color they picture is not n...

Colors in Context: A Pragmatic Neural Model for Grounded Language Understanding

We present a model of pragmatic referring expression interpretation in a...

Text2Shape: Generating Shapes from Natural Language by Learning Joint Embeddings

We present a method for generating colored 3D shapes from natural langua...

Modeling Color Terminology Across Thousands of Languages

There is an extensive history of scholarship into what constitutes a "ba...

Grounded Compositional Outputs for Adaptive Language Modeling

Language models have emerged as a central component across NLP, and a gr...

Fast Fourier Color Constancy

We present Fast Fourier Color Constancy (FFCC), a color constancy algori...

Code Repositories


Code for Learning to Generate Compositional Color Descriptions

view repo