Imitating Human Behaviour with Diffusion Models

By Tim Pearce, Tabish Rashid, Anssi Kanervisto, David Bignell, Mingfei Sun, Raluca Georgescu, Sergio Valcarcel Macua, Shan Zheng Tan, Ida Momennejad, Katja Hofmann and 1 more. 2023

323 citations recorded by the source corpus.

Diffusion models have emerged as powerful generative models in the text-to-image domain. This paper studies their application as observation-to-action models for imitating human behaviour in sequential environments. Human behaviour is stochastic and multimodal, with structured correlations between action dimensions. Meanwhile, standard modelling choices in behaviour cloning are limited in their expressiveness and may introduce bias into the cloned policy. We begin by pointing out the limitations of these choices. We then propose that diffusion models are an excellent fit for imitating human behaviour, since they learn an expressive distribution over the joint action space. We introduce several innovations to make diffusion models suitable for sequential environments; designing suitable architectures, investigating the role of guidance, and developing reliable sampling strategies. Experimentally, diffusion models closely match human demonstrations in a simulated robotic control task and a modern 3D gaming environment.

Subfields: Computer Vision and Pattern Recognition, Control and Systems Engineering, Generative Models

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Control and Systems Engineering

In Generative Models

Citations in the corpus

Diffusion policy: Visuomotor policy learning via action diffusion (2023, background): “[30] has conducted a complimentary analysis of diffusion-based policies in simulated environments.”
Is Conditional Generative Modeling all you need for Decision-Making? (2022, methodology): “Recent works have also used diffusion models to imitate human behavior (Pearce et al., 2023) and to parameterize policy in offline RL (Wang et al.”
IDQL: Implicit Q-Learning as an Actor-Critic Method with Diffusion Policies (2023, unknown): “Though, similar architectures for action modeling with diffusion have been proposed before [7, 36, 40].”
RDT-1B: a Diffusion Foundation Model for Bimanual Manipulation (2024, unknown): “While demonstrating generalization to new objects and tasks, they face issues with quantization errors and uncoordinated behaviors (Pearce et al., 2023) when applied to bimanual manipulation.”
CogACT: A Foundational Vision-Language-Action Model for Synergizing Cognition and Action in Robotic Manipulation (2024, unknown): “Recent studies [15, 50, 53] have introduced diffusion models as an innovative approach for modeling robotic actions.”
Diffusion Policies as an Expressive Policy Class for Offline Reinforcement Learning (2022, background): “Offline RL.”
3D Diffuser Actor: Policy Diffusion with 3D Scene Representations (2024, unknown): “Recent works use diffusion models for learning such state-conditioned action distributions for robot manipulation policies from demonstrations [54, 11, 60] and show they outperform deterministic or other alternatives, such as variational autoencoders [47], mixture of Gaussians † Equal contribution…”