CVPR Findings2026
A Simple Framework for Visual Navigation
Faith Johnson, Bryan Bo Cao, Shubham Jain, Ashwin Ashok, Kristin Dana
摘要
Visual navigation for robotics is inspired by the human ability to navigate environments using visual cues and memory, eliminating the need for detailed maps. In unseen, unmapped, or GPS-denied settings, traditional metric mapbased methods fall short, prompting a shift toward learningbased approaches with minimal exploration. In this work, we develop a hierarchical framework 1 that decomposes the navigation decision-making process into multiple levels.
Our method learns to select subgoals through a simple, transferable waypoint selection network. A key component of the approach is a latent-space memory module organized solely by visual similarity, as a proxy for distance. This alternative to graph-based topological representations proves sufficient for navigation tasks, providing a compact, lightweight, simple-to-train navigator that can find its way to the goal in novel locations. We show competitive results with a suite of SOTA methods in Habitat AI environments without using any odometry in training or inference. An additional contribution leverages the interpretability of the framework for interactive navigation. We consider the question: how much direction intervention/interaction is needed to achieve success in all trials? We demonstrate that even minimal human involvement can significantly enhance overall navigation performance.
问 Lune
几分钟读透这篇 Paper,不用再花一下午。
Lune 能读完全文,也能读引用它的顶会 Paper,讲清它的结论、证据和局限,每一点都注明出自哪篇 Paper。不用硬啃,也不靠猜。
脉络
- Object Goal Navigation using Goal-Oriented Semantic Exploration 引用了 Learning To Explore Using Active Neural SLAM
- Object Goal Navigation using Goal-Oriented Semantic Exploration 引用了 Habitat: A Platform for Embodied AI Research
- Learning To Explore Using Active Neural SLAM 引用了 Habitat: A Platform for Embodied AI Research
它引用的 Paper20
- Habitat: A Platform for Embodied AI ResearchManolis Savva, Jitendra Malik, Devi Parikh, Dhruv Batra 等ICCV 2019 · 1,863 次引用
- A Minimalist Approach to Offline Reinforcement LearningScott Fujimoto, Shixiang Shane GuNeurIPS 2021 · 1,292 次引用
- Object Goal Navigation using Goal-Oriented Semantic ExplorationDevendra Singh Chaplot, Dhiraj Gandhi, Abhinav Gupta, Ruslan SalakhutdinovNeurIPS 2020 · 857 次引用
- Learning To Explore Using Active Neural SLAMDevendra Singh Chaplot, Dhiraj Gandhi, Saurabh Gupta, Abhinav Gupta 等ICLR 2020 · 603 次引用
- Goal-Conditioned Reinforcement Learning with Imagined SubgoalsElliot Chane-Sane, Cordelia Schmid, Ivan LaptevICML 2021 · 183 次引用
相关 Paper
- Visual Graph Memory with Unsupervised Representation for Visual NavigationObin Kwon, Nuri Kim, Yunho Choi, Hwiyeon Yoo 等ICCV 2021 · 86 次引用
- Learning Navigational Visual Representations with Semantic Map SupervisionYicong Hong, Yang Zhou, Ruiyi Zhang, Franck Dernoncourt 等ICCV 2023 · 56 次引用
- Implicit Obstacle Map-driven Indoor Navigation Model for Robust Obstacle AvoidanceWei Xie, Haobo Jiang, Shuo Gu, Jin XieACM MM 2023 · 9 次引用
- The Surprising Effectiveness of Visual Odometry Techniques for Embodied PointGoal NavigationXiaoming Zhao, Harsh Agrawal, Dhruv Batra, Alexander G. SchwingICCV 2021 · 50 次引用
- AwareVLN: Reasoning with Self-awareness for Vision-Language NavigationWenxuan Guo, Xiuwei Xu, Yichen Liu, Xiangyu Li 等CVPR 2026 · 7 次引用
