CoDraw: Collaborative Drawing as a Testbed for Grounded Goal-driven Communication

Abstract

In this work, we propose a goal-driven collaborative task that combines language, perception, and action. Specifically, we develop a Collaborative image-Drawing game between two agents, called CoDraw. Our game is grounded in a virtual world that contains movable clip art objects. The game involves two players: a Teller and a Drawer. The Teller sees an abstract scene containing multiple clip art pieces in a semantically meaningful configuration, while the Drawer tries to reconstruct the scene on an empty canvas using available clip art pieces. The two players communicate with each other using natural language. We collect the CoDraw dataset of ~10K dialogs consisting of ~138K messages exchanged between human players. We define protocols and metrics to evaluate learned agents in this testbed, highlighting the need for a novel "crosstalk" evaluation condition which pairs agents trained independently on disjoint subsets of the training data. We present models for our task and benchmark them using both fully automated evaluation and by having them play the game live with humans.

arXiv commentACL 2019

References in corpus (19)

• Show, Attend and Tell: Neural Image Caption Generation with Visual Attention
• A Neural Conversational Model
• VQA: Visual Question Answering
• Learning to Communicate with Deep Multi-Agent Reinforcement Learning
• Show and Tell: Lessons learned from the 2015 MSCOCO Image Captioning Challenge
• Deep Reinforcement Learning for Dialogue Generation
• Exploring Models and Data for Image Question Answering
• Simple Baseline for Visual Question Answering
• Visual Genome: Connecting Language and Vision Using Crowdsourced Dense Image Annotations
• Are You Talking to a Machine? Dataset and Methods for Multilingual Image Question Answering
• Speaker-Follower Models for Vision-and-Language Navigation
• Image-Grounded Conversations: Multimodal Context for Natural Question and Response Generation
• Learning Cooperative Visual Dialog Agents with Deep Reinforcement Learning
• Learning through Dialogue Interactions by Asking Questions
• CLEVR: A Diagnostic Dataset for Compositional Language and Elementary Visual Reasoning
• End-to-end optimization of goal-driven and visually grounded dialogue systems
• It Takes Two to Tango: Towards Theory of AI's Mind
• GuessWhat?! Visual object discovery through multi-modal dialogue
• Navigational Instruction Generation as Inverse Reinforcement Learning with Neural Machine Translation

Cited by in corpus (7)

• Towards Ecologically Valid Research on Language User Interfaces
• Text as Neural Operator: Image Manipulation by Text Instruction
• StoryGAN: A Sequential Conditional GAN for Story Visualization
• Text2Scene: Generating Compositional Scenes from Textual Descriptions
• Image Generation from Layout
• Static and Animated 3D Scene Generation from Free-form Text Descriptions
• Chat-crowd: A Dialog-based Platform for Visual Layout Composition