Memory-Augmented Reinforcement Learning for Image-Goal Navigation
Open accessThe DOI matches its Crossref record; 51 of the 51 references it cites were found in a registry or an index.
Abstract
In this work, we present a memory-augmented approach for image-goal navigation. Earlier attempts, including RL-based and SLAM-based approaches have either shown poor generalization performance, or are heavily-reliant on pose/depth sensors. Our method is based on an attention-based end-to-end model that leverages an episodic memory to learn to navigate. First, we train a state-embedding network in a self-supervised fashion, and then use it to embed previously-visited states into the agent's memory. Our navigation policy takes advantage of this information through an attention mechanism. We validate our approach with extensive evaluations, and show that our model establishes a new state of the art on the challenging Gibson dataset. Furthermore, we achieve this impressive performance from RGB input alone, without access to additional information such as position or depth, in stark contrast to related work.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“A typical video understanding pipeline driven by long-term memory converts video frames into feature vectors, which can later be retrieved to support various tasks [11, 12, 13, 14, 15].”
From Human-inspired Perspectives: A Survey on AI Long-term Memory (2024) · cites this paper as [12]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“The aforementioned TDN methods [5], [8]-[13] address the problem of a single robot navigating to a target location in an unknown static environment. However, these methods do not consider dynamic obstacles (i.e., other robots) in their environments, leading to inaccuracies in spatial reasoning and degraded navigation performance [7].”
From NavFormer: A Transformer Architecture for Robot Target-Driven Navigation in Unknown and Dynamic Environments (2024) · cites this paper as [5]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“TDN in unknown and dynamic environments is a challenging problem as: 1) there are no global maps of the environment available, therefore, a robot needs to reason about the spatial layout of the environment based on its own partial observations to prevent deadlocks and redundant coverage [5]”
From NavFormer: A Transformer Architecture for Robot Target-Driven Navigation in Unknown and Dynamic Environments (2024) · cites this paper as [5]
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“To-date, existing robot TDN methods for unknown environments have mainly used: 1) deep reinforcement learning (DRL) [5], [8]-[12] or 2) imitation learning (IL) [13].”
From NavFormer: A Transformer Architecture for Robot Target-Driven Navigation in Unknown and Dynamic Environments (2024) · cites this paper as [5]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, L., & Polosukhin, I. (2017) · arXiv · doi:10.48550/arxiv.1706.03762 · Cited by 120,404 · cited 2×
“See [39] for more details about these sublayers.”
Exploiting Generative AI to Scale up Intelligent Tutoring Systems
Jakubův, J., Chvalovský, K., Goertzel, Z., Kaliszyk, C., Olšák, M., Piotrowski, B., Schulz, S., Suda, M., & Urban, J. (2023) · DROPS (Schloss Dagstuhl – Leibniz Center for Informatics) · doi:10.4230/lipics.itp.2023.19 · Cited by 76,658
Thrun, S. (2002) · Communications of the ACM · doi:10.1145/504729.504754 · Cited by 7,994 · cited 3×
“The task of navigation has been studied in the context of simultaneous localization and mapping (SLAM) in robotics [5].”
ORB-SLAM: A Versatile and Accurate Monocular SLAM System
Mur-Artal, R., Montiel, J. M. M., & Tardos, J. D. (2015) · IEEE Transactions on Robotics · doi:10.1109/tro.2015.2463671 · Cited by 6,778 · cited 1×
“Several SLAM methods comprise multiple hand-crafted modules to address strictly-defined problems in specific environments [12, 25, 30].”
ORB-SLAM: A Versatile and Accurate Monocular SLAM System
Mur-Artal, R., Montiel, J. M. M., & Tardos, J. D. (2015) · Zaguan (Universidad de Zaragoza) · Cited by 3,153
Target-driven visual navigation in indoor scenes using deep reinforcement learning
Zhu, Y., Mottaghi, R., Kolve, E., Lim, J. J., Gupta, A., Fei-Fei, L., & Farhadi, A. (2017) · 2017 IEEE International Conference on Robotics and Automation (ICRA) · doi:10.1109/icra.2017.7989381 · Cited by 1,533 · cited 3×
“Target-Driven RL [4]: A vanilla baseline trained with reinforcement learning, similar to our ablated variant.”
Habitat: A Platform for Embodied AI Research
Savva, M., Kadian, A., Maksymets, O., Zhao, Y., Wijmans, E., Jain, B., Straub, J., Liu, J., Koltun, V., Malik, J., et al. (2019) · 2019 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2019.00943 · Cited by 1,237 · cited 3×
“To run this baseline, we adapt the available code 3 to Habitat.”
Self-citationGibson Env: Real-World Perception for Embodied AgentsPreprint
Xia, F., Zamir, A., He, Z.-Y., Sax, A., Malik, J., & Savarese, S. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.48550/arxiv.1808.10654 · Cited by 1,076 · cited 6×
“With this approach we establish the new state of the art on the challenging image-goal navigation task on the Gibson dataset [8] by a large margin.”
BackgroundMethodInfluentialGibson Env: Real-World Perception for Embodied Agents
Xia, F., Zamir, A. R., He, Z., Sax, A., Malik, J., & Savarese, S. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00945 · Cited by 763
QT-Opt: Scalable Deep Reinforcement Learning for Vision-Based Robotic Manipulation
Kalashnikov, D., Irpan, A., Pastor, P., Ibarz, J., Herzog, A., Jang, E., Quillen, D., Holly, E., Kalakrishnan, M., Vanhoucke, V., et al. (2018) · arXiv · doi:10.48550/arxiv.1806.10293 · Cited by 697 · cited 1×
“Although data augmentation has been used in robotics applications [21], we show that it helps in planning setups such as navigation in Gibson [38].”
Method
Cited by
Memory in the Age of AI Agents
Hu, Y., Liu, S., Yue, Y., Zhang, G., Liu, B., Zhu, F., Lin, J., Guo, H., Dou, S., Xi, Z., et al. (2025) · arXiv · doi:10.48550/arxiv.2512.13564 · Cited by 389
CoWs on Pasture: Baselines and Benchmarks for Language-Driven Zero-Shot Object Navigation
Gadre, S. Y., Wortsman, M., Ilharco, G., Schmidt, L., & Song, S. (2023) · 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr52729.2023.02219 · Cited by 176
ZSON: Zero-Shot Object-Goal Navigation Using Multimodal Goal Embeddings
Majumdar, A., Aggarwal, G., Devnani, B., Hoffman, J., & Batra, D. (2022) · neural information processing systems · doi:10.52202/068431-2343 · Cited by 80
Self-citationAffordances from Human Videos as a Versatile Representation for Robotics
Bahl, S., Mendonca, R., Chen, L., Jain, U., & Pathak, D. (2023) · 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr52729.2023.01324 · Cited by 119
“Goal images are frequently used to specify rewards in RL [3, 34, 38, 74, 82, 83, 91, 121, 138].”
BackgroundOVRL-V2: A simple state-of-art baseline for ImageNav and ObjectNavPreprint
Yadav, K., Majumdar, A., Ramrakhya, R., Yokoyama, N., Baevski, A., Kira, Z., Maksymets, O., & Batra, D. (2023) · arXiv (Cornell University) · doi:10.48550/arxiv.2303.07798 · Cited by 107
“We perform I MAGE N AV experiments using a standard dataset released by [24].”
BackgroundMethodInfluentialSelf-citation
Metadata
- Authors
- Mezghan, L., Sukhbaatar, S., Lavril, T., Maksymets, O., Batra, D., Bojanowski, P., & Alahari, K.
- Affiliations
- Institut polytechnique de Grenoble (FR)
- Meta (United States) (US)
- Georgia Institute of Technology (US)
- Venue
- 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)
- Publisher
- IEEE
- Published
- 2022-10-23
- Pages
- 3316-3323
- ISSN
- 2153-0858 (ISSN-L)
- OpenAlex
- W3120006980
- Cited by
- 61
- References
- 48
- Citations by year
- 2026: 8 · 2025: 15 · 2024: 18 · 2023: 11 · 2022: 7
- FWCI
- 7.05
- Type
- conference paper
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- HAL (submittedVersion, other-oa)
- arXiv (Cornell University)
- Funders
- Agence Nationale de la Recherche (ANR-18-CE23-0011)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Reinforcement Learning in Robotics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Robotics and Sensor-Based Localization
- Physical Sciences › Engineering › Control and Systems Engineering › Robot Manipulation and Learning
- Keywords
- episodic memory, attention mechanism, robot navigation, generalization to unseen environments
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Memory-Augmented Reinforcement Learning for Image-Goal Navigation. https://app.cochat.ai/papers/10.1109/iros47612.2022.9981090
- The paper
- Mezghan, L., Sukhbaatar, S., Lavril, T., Maksymets, O., Batra, D., Bojanowski, P., & Alahari, K. (2022). Memory-Augmented Reinforcement Learning for Image-Goal Navigation. 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). https://doi.org/10.1109/iros47612.2022.9981090
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.