# Bayesian Disturbance Injection: Robust Imitation Learning of Flexible Policies for Robot Manipulation

arXiv: https://arxiv.org/abs/2211.03393 · HTML: https://mlsanity.com/paper/2211.03393/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- Placed in Imitation learning methods, one of 13 directions in Offline RL and imitation learning from data.
- Builds on 15 of its 57 references that are in the map.
- First posted on arXiv 7 Nov 2022.

## Placed in

- Offline RL and imitation learning from data / [Imitation learning methods](https://mlsanity.com/offlinerl/imitation-learning-methods/)

## Builds on (most cited first, 15 of 15)

- [End to End Learning for Self-Driving Cars](https://mlsanity.com/paper/1604.07316/) (2016)
- [An Algorithmic Perspective on Imitation Learning](https://mlsanity.com/paper/1811.06711/) (2018)
- [Learning Hand-Eye Coordination for Robotic Grasping with Deep Learning and Large-Scale Data Collection](https://mlsanity.com/paper/1603.02199/) (2016)
- [Deep Imitation Learning for Complex Manipulation Tasks from Virtual Reality Teleoperation](https://mlsanity.com/paper/1710.04615/) (2017)
- [DART: Noise Injection for Robust Imitation Learning](https://mlsanity.com/paper/1703.09327/) (2017)
- [Vision-Based Multi-Task Manipulation for Inexpensive Robots Using End-To-End Learning from Demonstration](https://mlsanity.com/paper/1707.02920/) (2017)
- [Learning from Suboptimal Demonstration via Self-Supervised Reward Regression](https://mlsanity.com/paper/2010.11723/) (2020)
- [Fastfood: Approximate Kernel Expansions in Loglinear Time](https://mlsanity.com/paper/1408.3060/) (2014)
- [A survey of robot learning from demonstrations for Human-Robot Collaboration](https://mlsanity.com/paper/1710.08789/) (2017)
- [Kernelized Movement Primitives](https://mlsanity.com/paper/1708.08638/) (2017)
- [Regression Planning Networks](https://mlsanity.com/paper/1909.13072/) (2019)
- [MPC-Inspired Neural Network Policies for Sequential Decision Making](https://mlsanity.com/paper/1802.05803/) (2018)
- [Learning a Multi-Modal Policy via Imitating Demonstrations with Mixed Behaviors](https://mlsanity.com/paper/1903.10304/) (2019)
- [Disturbance-Injected Robust Imitation Learning with Task Achievement](https://mlsanity.com/paper/2205.04195/) (2022)
- [Bayesian Disturbance Injection: Robust Imitation Learning of Flexible Policies](https://mlsanity.com/paper/2103.13623/) (2021)

## Built on by (most cited first, 8 of 8)

- [Leveraging Demonstrator-perceived Precision for Safe Interactive Imitation Learning of Clearance-limited Tasks](https://mlsanity.com/paper/2402.13466/) (2024)
- [Robust Instant Policy: Leveraging Student's t-Regression Model for Robust In-context Imitation Learning of Robot Manipulation](https://mlsanity.com/paper/2506.15157/) (2025)
- [Self-Augmented Robot Trajectory: Efficient Imitation Learning via Safe Self-augmentation with Demonstrator-annotated Precision](https://mlsanity.com/paper/2509.09893/) (2025)
- [Feasibility-aware Imitation Learning from Observations through a Hand-mounted Demonstration Interface](https://mlsanity.com/paper/2503.09018/) (2025)
- [Exploiting Policy Idling for Dexterous Manipulation](https://mlsanity.com/paper/2508.15669/) (2025)
- [Composite Gaussian Processes Flows for Learning Discontinuous Multimodal Policies](https://mlsanity.com/paper/2502.01913/) (2025)
- [CubeDAgger: Interactive Imitation Learning for Dynamic Systems with Efficient yet Low-risk Interaction](https://mlsanity.com/paper/2505.04897/) (2025)
- [Feasibility-aware Imitation Learning from Observation with Multimodal Feedback](https://mlsanity.com/paper/2602.15351/) (2026)

## Abstract (arXiv)

Humans demonstrate a variety of interesting behavioral characteristics when performing tasks, such as selecting between seemingly equivalent optimal actions, performing recovery actions when deviating from the optimal trajectory, or moderating actions in response to sensed risks. However, imitation learning, which attempts to teach robots to perform these same tasks from observations of human demonstrations, often fails to capture such behavior. Specifically, commonly used learning algorithms embody inherent contradictions between the learning assumptions (e.g., single optimal action) and actual human behavior (e.g., multiple optimal actions), thereby limiting robot generalizability, applicability, and demonstration feasibility. To address this, this paper proposes designing imitation learning algorithms with a focus on utilizing human behavioral characteristics, thereby embodying principles for capturing and exploiting actual demonstrator behavioral characteristics. This paper presents the first imitation learning framework, Bayesian Disturbance Injection (BDI), that typifies human behavioral characteristics by incorporating model flexibility, robustification, and risk sensitivity. Bayesian inference is used to learn flexible non-parametric multi-action policies, while simultaneously robustifying policies by injecting risk-sensitive disturbances to induce human recovery action and ensuring demonstration feasibility. Our method is evaluated through risk-sensitive simulations and real-robot experiments (e.g., table-sweep task, shaft-reach task and shaft-insertion task) using the UR5e 6-DOF robotic arm, to demonstrate the improved characterisation of behavior. Results show significant improvement in task performance, through improved flexibility, robustness as well as demonstration feasibility.