Pith. sign in

REVIEW 7 cited by

ZSON: Zero-Shot Object-Goal Navigation using Multimodal Goal Embeddings

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2206.12403 v2 pith:4FS6MOAC submitted 2022-06-24 cs.CV cs.LGcs.RO

Arjun Majumdar , Gunjan Aggarwal , Bhavika Devnani , Judy Hoffman , Dhruv Batra This is my paper · ORCID
classification cs.CVcs.LGcs.RO
keywords findagentsobjectnavsinknavigationapproachgoalmultimodal
verification ladder T0 review T1 audit T2 compute T3 formal

Signed reviews

No signed human review yet.

0 comments
read the original abstract

We present a scalable approach for learning open-world object-goal navigation (ObjectNav) -- the task of asking a virtual robot (agent) to find any instance of an object in an unexplored environment (e.g., "find a sink"). Our approach is entirely zero-shot -- i.e., it does not require ObjectNav rewards or demonstrations of any kind. Instead, we train on the image-goal navigation (ImageNav) task, in which agents find the location where a picture (i.e., goal image) was captured. Specifically, we encode goal images into a multimodal, semantic embedding space to enable training semantic-goal navigation (SemanticNav) agents at scale in unannotated 3D environments (e.g., HM3D). After training, SemanticNav agents can be instructed to find objects described in free-form natural language (e.g., "sink", "bathroom sink", etc.) by projecting language goals into the same multimodal, semantic embedding space. As a result, our approach enables open-world ObjectNav. We extensively evaluate our agents on three ObjectNav datasets (Gibson, HM3D, and MP3D) and observe absolute improvements in success of 4.2% - 20.0% over existing zero-shot methods. For reference, these gains are similar or better than the 5% improvement in success between the Habitat 2020 and 2021 ObjectNav challenge winners. In an open-world setting, we discover that our agents can generalize to compound instructions with a room explicitly mentioned (e.g., "Find a kitchen sink") and when the target room can be inferred (e.g., "Find a sink and a stove").

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 7 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. VTM-Nav: Harnessing Cross-Episode Experience for Object-Goal Navigation with Hierarchical Visual-Topological Memory

    cs.CV 2026-07 conditional novelty 6.0 of 10

    A hierarchical room-and-object memory that persists across independent ObjectNav episodes yields small success-rate gains, but most of the gain comes from within-episode memory rather than the cross-episode component.

  2. SplatSearch: Instance Image Goal Navigation for Mobile Robots using 3D Gaussian Splatting and Diffusion Models

    cs.RO 2025-11 conditional novelty 6.0 of 10

    SplatSearch combines sparse-view 3D Gaussian Splatting, multi-view diffusion inpainting, and semantic/visual frontier scoring to achieve viewpoint-invariant instance image-goal navigation in unknown environments.

  3. CARMA: Context-Aware Situational Grounding of Human-Robot Group Interactions by Combining Vision-Language Models with Object and Action Recognition

    cs.RO 2025-06 conditional novelty 6.0 of 10

    CARMA combines object detection, person tracking, action detection, and a vision-language model to produce instance-level actor-action-object triplets for human-robot group interactions, achieving up to 72% task succe...

  4. Efficient Policy Adaptation with Contrastive Prompt Ensemble for Embodied Agents

    cs.AI 2024-12 conditional novelty 6.0 of 10

    Contrastively learned visual prompts, combined through guided attention, improve zero-shot visual domain adaptation of embodied RL policies.

  5. History-Augmented Vision-Language Models for Frontier-Based Zero-Shot Object Navigation

    cs.RO 2025-06 conditional novelty 5.0 of 10

    Adding a history of recent actions to a vision-language model's prompt during frontier-based exploration yields 46% success and 24.8% SPL on HM3D zero-shot object navigation, on par with prior methods.

  6. GROOT-2: Weakly Supervised Multi-Modal Instruction Following Agents

    cs.AI 2024-12 conditional novelty 5.0 of 10

    A weakly supervised latent-variable agent improves multimodal instruction following by combining VAE self-imitating on unlabeled data with a likelihood-based alignment of labeled and video latents.

  7. Research on Navigation Methods Based on LLMs

    cs.RO 2025-04 reject novelty 3.0 of 10

    An LLM used as a central controller that picks modular navigation tools matches a conventional A* plus PID system in simulation, with no measured gain on navigation metrics.

Pith tools