DINO-WM: World Models on Pre-trained Visual Features enable Zero-shot Planning

• URL: http://arxiv.org/abs/2411.04983v2
• Date: Sat, 01 Feb 2025 02:40:49 GMT
• Title: DINO-WM: World Models on Pre-trained Visual Features enable Zero-shot Planning
• Abstract summary: DINO World Model (DINO-WM) is a new method to model visual dynamics without reconstructing the visual world.<n>We demonstrate that DINO-WM achieves zero-shot behavioral solutions at test time on six environments.
• Score: 38.749045283035365
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: The ability to predict future outcomes given control actions is fundamental for physical reasoning. However, such predictive models, often called world models, remains challenging to learn and are typically developed for task-specific solutions with online policy learning. To unlock world models' true potential, we argue that they should 1) be trainable on offline, pre-collected trajectories, 2) support test-time behavior optimization, and 3) facilitate task-agnostic reasoning. To this end, we present DINO World Model (DINO-WM), a new method to model visual dynamics without reconstructing the visual world. DINO-WM leverages spatial patch features pre-trained with DINOv2, enabling it to learn from offline behavioral trajectories by predicting future patch features. This allows DINO-WM to achieve observational goals through action sequence optimization, facilitating task-agnostic planning by treating goal features as prediction targets. We demonstrate that DINO-WM achieves zero-shot behavioral solutions at test time on six environments without expert demonstrations, reward modeling, or pre-learned inverse models, outperforming prior state-of-the-art work across diverse task families such as arbitrarily configured mazes, push manipulation with varied object shapes, and multi-particle scenarios.

Related papers

• HarmoWAM: Harmonizing Generalizable and Precise Manipulation via Adaptive World Action Models [58.191567345416836]
  World Action Models (WAMs) have emerged as a promising paradigm for robot control by modeling physical dynamics.<n>HarmoWAM fully leverages a world model to unify predictive and reactive control, enabling both general transit and precise manipulation.<n>We construct three training-unseen test environments across six real-world robotic tasks, covering variations in background, position, and object semantics.
  arXiv  Detail & Related papers  (2026-05-11T17:59:56Z)
• Closing the Train-Test Gap in World Models for Gradient-Based Planning [64.36544881136405]
  We propose improved methods for training world models that enable efficient gradient-based planning.<n>At test time, our approach outperforms or matches the classical gradient-free cross-entropy method.
  arXiv  Detail & Related papers  (2025-12-10T18:59:45Z)
• Learning Generalizable Visuomotor Policy through Dynamics-Alignment [13.655111993491674]
  Recent approaches leveraging video prediction models have shown promising results by learning rich representations from large-scale datasets.<n>We propose a Dynamics-Aligned Flow Matching Policy (DAP) that integrates dynamics prediction into policy learning.<n>Our method introduces a novel architecture where policy and dynamics models provide mutual corrective feedback during action generation, enabling self-correction and improved generalization.
  arXiv  Detail & Related papers  (2025-10-31T02:29:33Z)
• OMGPT: A Sequence Modeling Framework for Data-driven Operational Decision Making [5.419799294989289]
  We build a Generative Pre-trained Transformer (GPT) model to solve sequential decision making tasks.<n>We first propose a general sequence modeling framework to cover several operational decision making tasks.<n>We then train a transformer-based neural network model (OMGPT) as a natural and powerful architecture for sequential modeling.
  arXiv  Detail & Related papers  (2025-05-19T15:33:03Z)
• World Modeling Makes a Better Planner: Dual Preference Optimization for Embodied Task Planning [60.100794160682646]
  We propose a new learning framework that jointly optimize state prediction and action selection through preference learning. To automatically collect trajectories and stepwise preference data without human annotation, we introduce a tree search mechanism for extensive exploration via trial-and-error. Our method significantly outperforms existing methods and GPT-4o when applied to Qwen2-VL (7B), LLaVA-1.6 (7B), and LLaMA-3.2 (11B)
  arXiv  Detail & Related papers  (2025-03-13T15:49:56Z)
• Adaptive Planning with Generative Models under Uncertainty [20.922248169620783]
  Planning with generative models has emerged as an effective decision-making paradigm across a wide range of domains. While continuous replanning at each timestep might seem intuitive because it allows decisions to be made based on the most recent environmental observations, it results in substantial computational challenges. Our work addresses this challenge by introducing a simple adaptive planning policy that leverages the generative model's ability to predict long-horizon state trajectories.
  arXiv  Detail & Related papers  (2024-08-02T18:07:53Z)
• Self-Supervised Reinforcement Learning that Transfers using Random Features [41.00256493388967]
  We propose a self-supervised reinforcement learning method that enables the transfer of behaviors across tasks with different rewards. Our method is self-supervised in that it can be trained on offline datasets without reward labels, but can then be quickly deployed on new tasks.
  arXiv  Detail & Related papers  (2023-05-26T20:37:06Z)
• Predictive Experience Replay for Continual Visual Control and Forecasting [62.06183102362871]
  We present a new continual learning approach for visual dynamics modeling and explore its efficacy in visual control and forecasting. We first propose the mixture world model that learns task-specific dynamics priors with a mixture of Gaussians, and then introduce a new training strategy to overcome catastrophic forgetting. Our model remarkably outperforms the naive combinations of existing continual learning and visual RL algorithms on DeepMind Control and Meta-World benchmarks with continual visual control tasks.
  arXiv  Detail & Related papers  (2023-03-12T05:08:03Z)
• Reinforcement Learning with Action-Free Pre-Training from Videos [95.25074614579646]
  We introduce a framework that learns representations useful for understanding the dynamics via generative pre-training on videos. Our framework significantly improves both final performances and sample-efficiency of vision-based reinforcement learning.
  arXiv  Detail & Related papers  (2022-03-25T19:44:09Z)
• Procedure Planning in Instructional Videosvia Contextual Modeling and Model-based Policy Learning [114.1830997893756]
  This work focuses on learning a model to plan goal-directed actions in real-life videos. We propose novel algorithms to model human behaviors through Bayesian Inference and model-based Imitation Learning.
  arXiv  Detail & Related papers  (2021-10-05T01:06:53Z)
• Learning Dynamics Models for Model Predictive Agents [28.063080817465934]
  Model-Based Reinforcement Learning involves learning a textitdynamics model from data, and then using this model to optimise behaviour. This paper sets out to disambiguate the role of different design choices for learning dynamics models, by comparing their performance to planning with a ground-truth model.
  arXiv  Detail & Related papers  (2021-09-29T09:50:25Z)
• Model-Based Visual Planning with Self-Supervised Functional Distances [104.83979811803466]
  We present a self-supervised method for model-based visual goal reaching. Our approach learns entirely using offline, unlabeled data. We find that this approach substantially outperforms both model-free and model-based prior methods.
  arXiv  Detail & Related papers  (2020-12-30T23:59:09Z)
• Planning from Pixels using Inverse Dynamics Models [44.16528631970381]
  We propose a novel way to learn latent world models by learning to predict sequences of future actions conditioned on task completion. We evaluate our method on challenging visual goal completion tasks and show a substantial increase in performance compared to prior model-free approaches.
  arXiv  Detail & Related papers  (2020-12-04T06:07:36Z)
• Trajectory-wise Multiple Choice Learning for Dynamics Generalization in Reinforcement Learning [137.39196753245105]
  We present a new model-based reinforcement learning algorithm that learns a multi-headed dynamics model for dynamics generalization. We incorporate context learning, which encodes dynamics-specific information from past experiences into the context latent vector. Our method exhibits superior zero-shot generalization performance across a variety of control tasks, compared to state-of-the-art RL methods.
  arXiv  Detail & Related papers  (2020-10-26T03:20:42Z)
• Goal-Aware Prediction: Learning to Model What Matters [105.43098326577434]
  One of the fundamental challenges in using a learned forward dynamics model is the mismatch between the objective of the learned model and that of the downstream planner or policy. We propose to direct prediction towards task relevant information, enabling the model to be aware of the current task and encouraging it to only model relevant quantities of the state space. We find that our method more effectively models the relevant parts of the scene conditioned on the goal, and as a result outperforms standard task-agnostic dynamics models and model-free reinforcement learning.
  arXiv  Detail & Related papers  (2020-07-14T16:42:59Z)

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.