Think before Go: Hierarchical Reasoning for Image-goal Navigation

• URL: http://arxiv.org/abs/2604.17407v1
• Date: Sun, 19 Apr 2026 12:30:22 GMT
• Title: Think before Go: Hierarchical Reasoning for Image-goal Navigation
• Abstract summary: We propose a framework that decomposes image-goal navigation into high-level planning and low-level execution.<n>In high-level planning, a vision-language model is trained on a self-collected dataset to generate a short-horizon plan.<n>In low-level execution, an online reinforcement learning policy is utilized to decide actions conditioned on the short-horizon plan.
• Score: 42.91450373979155
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Image-goal navigation steers an agent to a target location specified by an image in unseen environments. Existing methods primarily handle this task by learning an end-to-end navigation policy, which compares the similarities of target and observation images and directly predicts the actions. However, when the target is distant or lies in another room, such methods fail to extract informative visual cues, leading the agent to wander around. Motivated by the human cognitive principle that deliberate, high-level reasoning guides fast, reactive execution in complex tasks, we propose Hierarchical Reasoning Navigation (HRNav), a framework that decomposes image-goal navigation into high-level planning and low-level execution. In high-level planning, a vision-language model is trained on a self-collected dataset to generate a short-horizon plan, such as whether the agent should walk through the door or down the hallway. This downgrades the difficulty of the long-horizon task, making it more amenable to the execution part. In low-level execution, an online reinforcement learning policy is utilized to decide actions conditioned on the short-horizon plan. We also devise a novel Wandering Suppression Penalty (WSP) to further reduce the wandering problem. Together, these components form a hierarchical framework for Image-Goal Navigation. Extensive experiments in both simulation and real-world environments demonstrate the superiority of our method.

Related papers

• Flying to Image-Specified Objects: 3D Quadrotor Navigation via Cross-Graph Memory and Viewpoint Planning [6.546667407424349]
  Instance-Specific Image-Goal Navigation (InstanceImageNav) requires a robot to navigate toward the exact object instance depicted in a query image.<n>We propose a hierarchical navigation framework for quadrotor InstanceImageNav that separates high-level decision making from low-level motion execution.
  arXiv  Detail & Related papers  (2026-06-29T07:54:24Z)
• Embodied Instruction Following in Unknown Environments [64.57388036567461]
  We propose an embodied instruction following (EIF) method for complex tasks in the unknown environment.<n>We build a hierarchical embodied instruction following framework including the high-level task planner and the low-level exploration controller.<n>For the task planner, we generate the feasible step-by-step plans for human goal accomplishment according to the task completion process and the known visual clues.
  arXiv  Detail & Related papers  (2024-06-17T17:55:40Z)
• Transformers for Image-Goal Navigation [0.0]
  We present a generative Transformer based model that jointly models image goals, camera observations and the robot's past actions to predict future actions. Our model demonstrates capability in capturing and associating visual information across long time horizons, helping in effective navigation.
  arXiv  Detail & Related papers  (2024-05-23T03:01:32Z)
• Instance-aware Exploration-Verification-Exploitation for Instance ImageGoal Navigation [88.84058353659107]
  Instance ImageGoal Navigation (IIN) aims to navigate to a specified object depicted by a goal image in an unexplored environment. We propose a new modular navigation framework named Instance-aware Exploration-Verification-Exploitation (IEVE) for instance-level image goal navigation. Our method surpasses previous state-of-the-art work, with a classical segmentation model (0.684 vs. 0.561 success) or a robust model (0.702 vs. 0.561 success)
  arXiv  Detail & Related papers  (2024-02-25T07:59:10Z)
• How To Not Train Your Dragon: Training-free Embodied Object Goal Navigation with Semantic Frontiers [94.46825166907831]
  We present a training-free solution to tackle the object goal navigation problem in Embodied AI. Our method builds a structured scene representation based on the classic visual simultaneous localization and mapping (V-SLAM) framework. Our method propagates semantics on the scene graphs based on language priors and scene statistics to introduce semantic knowledge to the geometric frontiers.
  arXiv  Detail & Related papers  (2023-05-26T13:38:33Z)
• Long-HOT: A Modular Hierarchical Approach for Long-Horizon Object Transport [83.06265788137443]
  We address key challenges in long-horizon embodied exploration and navigation by proposing a new object transport task and a novel modular framework for temporally extended navigation. Our first contribution is the design of a novel Long-HOT environment focused on deep exploration and long-horizon planning. We propose a modular hierarchical transport policy (HTP) that builds a topological graph of the scene to perform exploration with the help of weighted frontiers.
  arXiv  Detail & Related papers  (2022-10-28T05:30:49Z)
• Deep Hierarchical Planning from Pixels [86.14687388689204]
  Director is a method for learning hierarchical behaviors directly from pixels by planning inside the latent space of a learned world model. Despite operating in latent space, the decisions are interpretable because the world model can decode goals into images for visualization. Director also learns successful behaviors across a wide range of environments, including visual control, Atari games, and DMLab levels.
  arXiv  Detail & Related papers  (2022-06-08T18:20:15Z)
• Are We There Yet? Learning to Localize in Embodied Instruction Following [1.7300690315775575]
  Action Learning From Realistic Environments and Directives (ALFRED) is a recently proposed benchmark for this problem. Key challenges for this task include localizing target locations and navigating to them through visual inputs. We augment the agent's field of view during navigation subgoals with multiple viewing angles, and train the agent to predict its relative spatial relation to the target location at each timestep.
  arXiv  Detail & Related papers  (2021-01-09T21:49:41Z)

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.