Mapping Instructions and Visual Observations to Actions with Reinforcement Learning
This addresses the challenge of integrating linguistic and visual input for autonomous agents, but it is incremental as it builds on existing RL methods with reward shaping.
The paper tackles the problem of directly mapping raw visual observations and text instructions to actions for instruction execution, using a single neural network trained with reinforcement learning and reward shaping, and shows significant improvements over supervised learning and common RL variants in a simulated environment.
We propose to directly map raw visual observations and text input to actions for instruction execution. While existing approaches assume access to structured environment representations or use a pipeline of separately trained models, we learn a single model to jointly reason about linguistic and visual input. We use reinforcement learning in a contextual bandit setting to train a neural network agent. To guide the agent's exploration, we use reward shaping with different forms of supervision. Our approach does not require intermediate representations, planning procedures, or training different models. We evaluate in a simulated environment, and show significant improvements over supervised learning and common reinforcement learning variants.