Text2Action: Generative Adversarial Synthesis from Language to Action
This work addresses the challenge of enabling robots or virtual agents to perform actions based on textual commands, representing an incremental advancement in language-to-action synthesis.
The paper tackles the problem of generating human action sequences from natural language descriptions, using a generative adversarial network based on a sequence-to-sequence model trained on 29,770 action-sentence pairs from MSR-VTT, and demonstrates that it can produce diverse, human-like actions transferable to a Baxter robot.
In this paper, we propose a generative model which learns the relationship between language and human action in order to generate a human action sequence given a sentence describing human behavior. The proposed generative model is a generative adversarial network (GAN), which is based on the sequence to sequence (SEQ2SEQ) model. Using the proposed generative network, we can synthesize various actions for a robot or a virtual agent using a text encoder recurrent neural network (RNN) and an action decoder RNN. The proposed generative network is trained from 29,770 pairs of actions and sentence annotations extracted from MSR-Video-to-Text (MSR-VTT), a large-scale video dataset. We demonstrate that the network can generate human-like actions which can be transferred to a Baxter robot, such that the robot performs an action based on a provided sentence. Results show that the proposed generative network correctly models the relationship between language and action and can generate a diverse set of actions from the same sentence.