# Retrospective Open-Vocabulary Memory for Long-Term Object Search

arXiv: https://arxiv.org/abs/2610.00330 · HTML: https://mlsanity.com/paper/2610.00330/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 25 of its 29 references that are in the map.
- First posted on arXiv 29 Sep 2026.

## Builds on (most cited first, 25 of 25)

- [Sigmoid Loss for Language Image Pre-Training](https://mlsanity.com/paper/2303.15343/) (2023)
- [Habitat: A Platform for Embodied AI Research](https://mlsanity.com/paper/1904.01201/) (2019)
- [Google Scanned Objects: A High-Quality Dataset of 3D Scanned Household Items](https://mlsanity.com/paper/2204.11918/) (2022)
- [Habitat-Matterport 3D Dataset (HM3D): 1000 Large-scale 3D Environments for Embodied AI](https://mlsanity.com/paper/2109.08238/) (2021)
- [On Evaluation of Embodied Navigation Agents](https://mlsanity.com/paper/1807.06757/) (2018)
- [Object Goal Navigation using Goal-Oriented Semantic Exploration](https://mlsanity.com/paper/2007.00643/) (2020)
- [ProcTHOR: Large-Scale Embodied AI Using Procedural Generation](https://mlsanity.com/paper/2206.06994/) (2022)
- [Visual Language Maps for Robot Navigation](https://mlsanity.com/paper/2210.05714/) (2022)
- [ConceptGraphs: Open-Vocabulary 3D Scene Graphs for Perception and Planning](https://mlsanity.com/paper/2309.16650/) (2023)
- [Scaling Open-Vocabulary Object Detection](https://mlsanity.com/paper/2306.09683/) (2023)
- [ObjectNav Revisited: On Evaluation of Embodied Agents Navigating to Objects](https://mlsanity.com/paper/2006.13171/) (2020)
- [RIO: 3D Object Instance Re-Localization in Changing Indoor Environments](https://mlsanity.com/paper/1908.06109/) (2019)
- [PONI: Potential Functions for ObjectGoal Navigation with Interaction-free Learning](https://mlsanity.com/paper/2201.10029/) (2022)
- [Habitat-Matterport 3D Semantics Dataset](https://mlsanity.com/paper/2210.05633/) (2022)
- [GOAT-Bench: A Benchmark for Multi-Modal Lifelong Navigation](https://mlsanity.com/paper/2404.06609/) (2024)
- [Dynamic Open-Vocabulary 3D Scene Graphs for Long-term Language-Guided Mobile Manipulation](https://mlsanity.com/paper/2410.11989/) (2024)
- [DynaMem: Online Dynamic Spatio-Semantic Memory for Open World Mobile Manipulation](https://mlsanity.com/paper/2411.04999/) (2024)
- [Describe Anything Anywhere At Any Moment](https://mlsanity.com/paper/2512.00565/) (2025)
- [Probable Object Location (POLo) Score Estimation for Efficient Object Goal Navigation](https://mlsanity.com/paper/2311.07992/) (2023)
- [Predictive Spatio-Temporal Scene Graphs for Semi-Static Scenes](https://mlsanity.com/paper/2605.00121/) (2026)
- [Searching in Space and Time: Unified Memory-Action Loops for Open-World Object Retrieval](https://mlsanity.com/paper/2511.14004/) (2025)
- [Perpetua: Multi-Hypothesis Persistence Modeling for Semi-Static Environments](https://mlsanity.com/paper/2507.18808/) (2025)
- [Change-Robust Online Topological Memory for Long-Term Relocalization and Semantic Navigation](https://mlsanity.com/paper/2605.02227/) (2026)
- [RAVEN: Long-Horizon Reasoning & Navigation with a Visuo-Spatio-Temporal Memory](https://mlsanity.com/paper/2606.25206/) (2026)
- [ObsGraph: Hierarchical Observation Representation for Embodied Reasoning and Exploration](https://mlsanity.com/paper/2606.24068/) (2026)

## Abstract (arXiv)

Long-term object search requires learning where objects usually appear from repeated but uneven observations of a changing environment. We formulate retrospective open-vocabulary memory as probabilistic inference from censored observations, where the key idea is to reason with evidence per opportunity: a detection or non-detection should influence belief only in proportion to the robot's opportunity to observe the corresponding location. We introduce ECROM, which uses this principle to estimate long-term prevalence for concepts specified only at query time and converts the resulting belief directly into an active-search prior. To evaluate this problem, we introduce a controlled long-term benchmark in ten HM3D homes that independently varies object placement and observation opportunity across repeated traversals. ECROM improves support-level AP on held-out queries by 4.5 points and search SPL by 4.2 points over the strongest competing memory in each metric. The benchmark, dataset, and code will be open-sourced. Project page: https://jiaming.im/ecrom/