Objects matter: object-centric world models improve reinforcement learning in visually complex environments
- URL: http://arxiv.org/abs/2501.16443v1
- Date: Mon, 27 Jan 2025 19:07:06 GMT
- Title: Objects matter: object-centric world models improve reinforcement learning in visually complex environments
- Authors: Weipu Zhang, Adam Jelley, Trevor McInroe, Amos Storkey,
- Abstract summary: We propose an object-centric reinforcement learning pipeline for visually complex games.
We show how this pipeline can overcome the limitations of traditional model-based reinforcement learning.
- Score: 2.2532647717990937
- License:
- Abstract: Deep reinforcement learning has achieved remarkable success in learning control policies from pixels across a wide range of tasks, yet its application remains hindered by low sample efficiency, requiring significantly more environment interactions than humans to reach comparable performance. Model-based reinforcement learning (MBRL) offers a solution by leveraging learnt world models to generate simulated experience, thereby improving sample efficiency. However, in visually complex environments, small or dynamic elements can be critical for decision-making. Yet, traditional MBRL methods in pixel-based environments typically rely on auto-encoding with an $L_2$ loss, which is dominated by large areas and often fails to capture decision-relevant details. To address these limitations, we propose an object-centric MBRL pipeline, which integrates recent advances in computer vision to allow agents to focus on key decision-related elements. Our approach consists of four main steps: (1) annotating key objects related to rewards and goals with segmentation masks, (2) extracting object features using a pre-trained, frozen foundation vision model, (3) incorporating these object features with the raw observations to predict environmental dynamics, and (4) training the policy using imagined trajectories generated by this object-centric world model. Building on the efficient MBRL algorithm STORM, we call this pipeline OC-STORM. We demonstrate OC-STORM's practical value in overcoming the limitations of conventional MBRL approaches on both Atari games and the visually complex game Hollow Knight.
Related papers
- Exploring the limits of Hierarchical World Models in Reinforcement Learning [0.7499722271664147]
We describe a novel HMBRL framework and evaluate it thoroughly.
We construct hierarchical world models that simulate environment dynamics at various levels of temporal abstraction.
Unlike most goal-conditioned H(MB)RL approaches, it also leads to comparatively low dimensional abstract actions.
arXiv Detail & Related papers (2024-06-01T16:29:03Z) - Learning Latent Dynamic Robust Representations for World Models [9.806852421730165]
Visual Model-Based Reinforcement Learning (MBL) promises to agent's knowledge about the underlying dynamics of the environment.
Top-temporal agents such as Dreamer often struggle with visual pixel-based inputs in the presence of irrelevant noise in the observation space.
We apply a-temporal masking strategy, combined with latent reconstruction, to capture endogenous task-specific aspects of the environment for world models.
arXiv Detail & Related papers (2024-05-10T06:28:42Z) - Graphical Object-Centric Actor-Critic [55.2480439325792]
We propose a novel object-centric reinforcement learning algorithm combining actor-critic and model-based approaches.
We use a transformer encoder to extract object representations and graph neural networks to approximate the dynamics of an environment.
Our algorithm performs better in a visually complex 3D robotic environment and a 2D environment with compositional structure than the state-of-the-art model-free actor-critic algorithm.
arXiv Detail & Related papers (2023-10-26T06:05:12Z) - HarmonyDream: Task Harmonization Inside World Models [93.07314830304193]
Model-based reinforcement learning (MBRL) holds the promise of sample-efficient learning.
We propose a simple yet effective approach, HarmonyDream, which automatically adjusts loss coefficients to maintain task harmonization.
arXiv Detail & Related papers (2023-09-30T11:38:13Z) - Weakly-supervised Contrastive Learning for Unsupervised Object Discovery [52.696041556640516]
Unsupervised object discovery is promising due to its ability to discover objects in a generic manner.
We design a semantic-guided self-supervised learning model to extract high-level semantic features from images.
We introduce Principal Component Analysis (PCA) to localize object regions.
arXiv Detail & Related papers (2023-07-07T04:03:48Z) - Bridging the Gap to Real-World Object-Centric Learning [66.55867830853803]
We show that reconstructing features from models trained in a self-supervised manner is a sufficient training signal for object-centric representations to arise in a fully unsupervised way.
Our approach, DINOSAUR, significantly out-performs existing object-centric learning models on simulated data.
arXiv Detail & Related papers (2022-09-29T15:24:47Z) - Multitask Adaptation by Retrospective Exploration with Learned World
Models [77.34726150561087]
We propose a meta-learned addressing model called RAMa that provides training samples for the MBRL agent taken from task-agnostic storage.
The model is trained to maximize the expected agent's performance by selecting promising trajectories solving prior tasks from the storage.
arXiv Detail & Related papers (2021-10-25T20:02:57Z) - Model-Advantage Optimization for Model-Based Reinforcement Learning [41.13567626667456]
Model-based Reinforcement Learning (MBRL) algorithms have been traditionally designed with the goal of learning accurate dynamics of the environment.
Value-aware model learning, an alternative model-learning paradigm to maximum likelihood, proposes to inform model-learning through the value function of the learnt policy.
We propose a novel value-aware objective that is an upper bound on the absolute performance difference of a policy across two models.
arXiv Detail & Related papers (2021-06-26T20:01:28Z) - Reinforcement Learning for Sparse-Reward Object-Interaction Tasks in a
First-person Simulated 3D Environment [73.9469267445146]
First-person object-interaction tasks in high-fidelity, 3D, simulated environments such as the AI2Thor pose significant sample-efficiency challenges for reinforcement learning agents.
We show that one can learn object-interaction tasks from scratch without supervision by learning an attentive object-model as an auxiliary task.
arXiv Detail & Related papers (2020-10-28T19:27:26Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.