@inproceedings{chung-yoon-2026-improving,
title = "Improving Allocentric Spatial Reasoning of {VLM}s via Perspective-conditional Prompt-based Contrastive Decoding",
author = "Chung, Hyejin and
Yoon, Hyunsoo",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.1037/",
doi = "10.18653/2026.findings-emnlp.1037",
pages = "21917--21935",
ISBN = "979-8-89176-462-0",
abstract = "Vision-Language models (VLMs) have demonstrated impressive capabilities across diverse domains, yet they fail significantly in allocentric spatial reasoning, which requires perspective-taking from object-centric viewpoints rather than the camera{'}s perspective. This capability is essential for real-world applications including robotics, autonomous driving, and embodied AI. In this work, we identify this limitation as a prevalent camera viewpoint bias and propose Perspective-Conditional Prompt-based Contrastive Decoding (PCD), a training-free method that leverages VLMs' capabilities while exploiting this performance gap. By contrasting responses to allocentric versus egocentric framings of identical spatial configurations, PCD explicitly suppresses camera viewpoint bias without requiring fine-tuning or external foundation models. Evaluating on allocentric benchmarks including 3DSRBench, OmniSpatial, ViewSpatial-Bench, and SpatialMQA, we demonstrate that PCD improves accuracy by up to 32.3 points on Qwen2.5-VL-7B, with consistent gains observed on recent, larger-scale VLMs. Our approach offers substantial computational advantages, achieving 8.65{\texttimes} faster inference and a 37.4{\%} memory reduction compared to multi-model pipeline framework."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="chung-yoon-2026-improving">
<titleInfo>
<title>Improving Allocentric Spatial Reasoning of VLMs via Perspective-conditional Prompt-based Contrastive Decoding</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hyejin</namePart>
<namePart type="family">Chung</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hyunsoo</namePart>
<namePart type="family">Yoon</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-462-0</identifier>
</relatedItem>
<abstract>Vision-Language models (VLMs) have demonstrated impressive capabilities across diverse domains, yet they fail significantly in allocentric spatial reasoning, which requires perspective-taking from object-centric viewpoints rather than the camera’s perspective. This capability is essential for real-world applications including robotics, autonomous driving, and embodied AI. In this work, we identify this limitation as a prevalent camera viewpoint bias and propose Perspective-Conditional Prompt-based Contrastive Decoding (PCD), a training-free method that leverages VLMs’ capabilities while exploiting this performance gap. By contrasting responses to allocentric versus egocentric framings of identical spatial configurations, PCD explicitly suppresses camera viewpoint bias without requiring fine-tuning or external foundation models. Evaluating on allocentric benchmarks including 3DSRBench, OmniSpatial, ViewSpatial-Bench, and SpatialMQA, we demonstrate that PCD improves accuracy by up to 32.3 points on Qwen2.5-VL-7B, with consistent gains observed on recent, larger-scale VLMs. Our approach offers substantial computational advantages, achieving 8.65× faster inference and a 37.4% memory reduction compared to multi-model pipeline framework.</abstract>
<identifier type="citekey">chung-yoon-2026-improving</identifier>
<identifier type="doi">10.18653/2026.findings-emnlp.1037</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.1037/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>21917</start>
<end>21935</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Improving Allocentric Spatial Reasoning of VLMs via Perspective-conditional Prompt-based Contrastive Decoding
%A Chung, Hyejin
%A Yoon, Hyunsoo
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F chung-yoon-2026-improving
%X Vision-Language models (VLMs) have demonstrated impressive capabilities across diverse domains, yet they fail significantly in allocentric spatial reasoning, which requires perspective-taking from object-centric viewpoints rather than the camera’s perspective. This capability is essential for real-world applications including robotics, autonomous driving, and embodied AI. In this work, we identify this limitation as a prevalent camera viewpoint bias and propose Perspective-Conditional Prompt-based Contrastive Decoding (PCD), a training-free method that leverages VLMs’ capabilities while exploiting this performance gap. By contrasting responses to allocentric versus egocentric framings of identical spatial configurations, PCD explicitly suppresses camera viewpoint bias without requiring fine-tuning or external foundation models. Evaluating on allocentric benchmarks including 3DSRBench, OmniSpatial, ViewSpatial-Bench, and SpatialMQA, we demonstrate that PCD improves accuracy by up to 32.3 points on Qwen2.5-VL-7B, with consistent gains observed on recent, larger-scale VLMs. Our approach offers substantial computational advantages, achieving 8.65× faster inference and a 37.4% memory reduction compared to multi-model pipeline framework.
%R 10.18653/2026.findings-emnlp.1037
%U https://aclanthology.org/2026.findings-emnlp.1037/
%U https://doi.org/10.18653/2026.findings-emnlp.1037
%P 21917-21935
Markdown (Informal)
[Improving Allocentric Spatial Reasoning of VLMs via Perspective-conditional Prompt-based Contrastive Decoding](https://aclanthology.org/2026.findings-emnlp.1037/) (Chung & Yoon, Findings 2026)
ACL