# Deep Active Inference for Pixel-Based Discrete Control: Evaluation on the Car Racing Problem

arXiv: https://arxiv.org/abs/2109.04155 · HTML: https://mlsanity.com/paper/2109.04155/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 16 of its 27 references that are in the map.
- First posted on arXiv 9 Sep 2021.

## Builds on (most cited first, 16 of 16)

- [World Models](https://mlsanity.com/paper/1803.10122/) (2018)
- [Active inference on discrete state-spaces: a synthesis](https://mlsanity.com/paper/2001.07203/) (2020)
- [Deep active inference agents using Monte-Carlo methods](https://mlsanity.com/paper/2006.04176/) (2020)
- [Weight Agnostic Neural Networks](https://mlsanity.com/paper/1906.04358/) (2019)
- [Deep Active Inference as Variational Policy Gradients](https://mlsanity.com/paper/1907.03876/) (2019)
- [Reinforcement Learning through Active Inference](https://mlsanity.com/paper/2002.12636/) (2020)
- [Deep Active Inference](https://mlsanity.com/paper/1709.02341/) (2017)
- [Scaling active inference](https://mlsanity.com/paper/1911.10601/) (2019)
- [End-to-End Pixel-Based Deep Active Inference for Body Perception and Action](https://mlsanity.com/paper/2001.05847/) (2019)
- [Deep Neuroevolution of Recurrent and Discrete World Models](https://mlsanity.com/paper/1906.08857/) (2019)
- [Multimodal VAE Active Inference Controller](https://mlsanity.com/paper/2103.04412/) (2021)
- [Deep Active Inference for Partially Observable MDPs](https://mlsanity.com/paper/2009.03622/) (2020)
- [Bayesian policy selection using active inference](https://mlsanity.com/paper/1904.08149/) (2019)
- [Neuroscience-inspired perception-action in robotics: applying active inference for state estimation, control and self-perception](https://mlsanity.com/paper/2105.04261/) (2021)
- [Online reinforcement learning with sparse rewards through an active inference capsule](https://mlsanity.com/paper/2106.02390/) (2021)
- [Deep Reinforcement Learning with Mixed Convolutional Network](https://mlsanity.com/paper/2010.00717/) (2020)

## Built on by (most cited first, 2 of 2)

- [Active Inference in Robotics and Artificial Agents: Survey and Challenges](https://mlsanity.com/paper/2112.01871/) (2021)
- [R-AIF: Solving Sparse-Reward Robotic Tasks from Pixels with Active Inference and World Models](https://mlsanity.com/paper/2409.14216/) (2024)

## Abstract (arXiv)

Despite the potential of active inference for visual-based control, learning the model and the preferences (priors) while interacting with the environment is challenging. Here, we study the performance of a deep active inference (dAIF) agent on OpenAI's car racing benchmark, where there is no access to the car's state. The agent learns to encode the world's state from high-dimensional input through unsupervised representation learning. State inference and control are learned end-to-end by optimizing the expected free energy. Results show that our model achieves comparable performance to deep Q-learning. However, vanilla dAIF does not reach state-of-the-art performance compared to other world model approaches. Hence, we discuss the current model implementation's limitations and potential architectures to overcome them.