@inproceedings{gaafar-saqr-2026-ensar,
title = "Ensar at {G}round{LM} 2026 Shared Tasks: Symbolic Scene Serialisation for View-Level Evidence Identification",
author = "Gaafar, Mostafa and
Saqr, Omar Mokhtar Gaber",
editor = "Wang, Yimu and
Choi, Yee Man and
Wu, Di and
Ouyang, Siqi and
Azadani, Mozhgan Nasr and
Fung, Yi R. (May)",
booktitle = "Proceedings of the 1st Workshop on Grounding Language Models: Learning Faithfully and Efficiently ({G}round{LM} 2026)",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.groundlm-1.90/",
pages = "1214--1220",
ISBN = "979-8-89176-478-1",
abstract = "An autonomous vehicle has the privilege of abundant data drawn from many sources, but abundance raises a question of provenance: when we care about one specific event in the scene, which source actually carries the evi dence for it? Identifying that source matters be yond attribution, because a system that weighs every channel at once lets the irrelevant ones di lute the decision, and more input is not the same as more information. GoldenViewVQA poses this question in a controlled form, asking a sys tem to name which of six synchronised vehicle cameras carries the evidence for a question, and to answer that question. We introduce a novel two-stage system built on symbolic scene serialisation. We sweep all six views with a de tector fine-tuned on nuScenes, back-project ev ery detection onto the ground plane, associate detections across cameras into single objects that carry an explicit view-set, and render the scene as text. A language model then names the supporting view and the answer, reading either the serialisation alone (System 1) or the serialisation together with the six raw images (System 2). The pixels help the system locate evidence, not answer questions: System 2 lifts view macro-accuracy to 0.68, against 0.38 for System 1 and 0.14 for an always-CAM{\_}FRONT baseline. Our best run reaches 0.78 joint accu racy, 0.83 view accuracy and 0.65 view macro accuracy on the official test split. Answer accu racy already reaches 0.95 with no image at all: option phrasing alone largely solves the answer head, leaving nearly all the measurable signal in the view head."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="gaafar-saqr-2026-ensar">
<titleInfo>
<title>Ensar at GroundLM 2026 Shared Tasks: Symbolic Scene Serialisation for View-Level Evidence Identification</title>
</titleInfo>
<name type="personal">
<namePart type="given">Mostafa</namePart>
<namePart type="family">Gaafar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Omar</namePart>
<namePart type="given">Mokhtar</namePart>
<namePart type="given">Gaber</namePart>
<namePart type="family">Saqr</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>Proceedings of the 1st Workshop on Grounding Language Models: Learning Faithfully and Efficiently (GroundLM 2026)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yimu</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yee</namePart>
<namePart type="given">Man</namePart>
<namePart type="family">Choi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Di</namePart>
<namePart type="family">Wu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Siqi</namePart>
<namePart type="family">Ouyang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mozhgan</namePart>
<namePart type="given">Nasr</namePart>
<namePart type="family">Azadani</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yi</namePart>
<namePart type="given">R</namePart>
<namePart type="given">(May)</namePart>
<namePart type="family">Fung</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">979-8-89176-478-1</identifier>
</relatedItem>
<abstract>An autonomous vehicle has the privilege of abundant data drawn from many sources, but abundance raises a question of provenance: when we care about one specific event in the scene, which source actually carries the evi dence for it? Identifying that source matters be yond attribution, because a system that weighs every channel at once lets the irrelevant ones di lute the decision, and more input is not the same as more information. GoldenViewVQA poses this question in a controlled form, asking a sys tem to name which of six synchronised vehicle cameras carries the evidence for a question, and to answer that question. We introduce a novel two-stage system built on symbolic scene serialisation. We sweep all six views with a de tector fine-tuned on nuScenes, back-project ev ery detection onto the ground plane, associate detections across cameras into single objects that carry an explicit view-set, and render the scene as text. A language model then names the supporting view and the answer, reading either the serialisation alone (System 1) or the serialisation together with the six raw images (System 2). The pixels help the system locate evidence, not answer questions: System 2 lifts view macro-accuracy to 0.68, against 0.38 for System 1 and 0.14 for an always-CAM_FRONT baseline. Our best run reaches 0.78 joint accu racy, 0.83 view accuracy and 0.65 view macro accuracy on the official test split. Answer accu racy already reaches 0.95 with no image at all: option phrasing alone largely solves the answer head, leaving nearly all the measurable signal in the view head.</abstract>
<identifier type="citekey">gaafar-saqr-2026-ensar</identifier>
<location>
<url>https://aclanthology.org/2026.groundlm-1.90/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>1214</start>
<end>1220</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Ensar at GroundLM 2026 Shared Tasks: Symbolic Scene Serialisation for View-Level Evidence Identification
%A Gaafar, Mostafa
%A Saqr, Omar Mokhtar Gaber
%Y Wang, Yimu
%Y Choi, Yee Man
%Y Wu, Di
%Y Ouyang, Siqi
%Y Azadani, Mozhgan Nasr
%Y Fung, Yi R. (May)
%S Proceedings of the 1st Workshop on Grounding Language Models: Learning Faithfully and Efficiently (GroundLM 2026)
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-478-1
%F gaafar-saqr-2026-ensar
%X An autonomous vehicle has the privilege of abundant data drawn from many sources, but abundance raises a question of provenance: when we care about one specific event in the scene, which source actually carries the evi dence for it? Identifying that source matters be yond attribution, because a system that weighs every channel at once lets the irrelevant ones di lute the decision, and more input is not the same as more information. GoldenViewVQA poses this question in a controlled form, asking a sys tem to name which of six synchronised vehicle cameras carries the evidence for a question, and to answer that question. We introduce a novel two-stage system built on symbolic scene serialisation. We sweep all six views with a de tector fine-tuned on nuScenes, back-project ev ery detection onto the ground plane, associate detections across cameras into single objects that carry an explicit view-set, and render the scene as text. A language model then names the supporting view and the answer, reading either the serialisation alone (System 1) or the serialisation together with the six raw images (System 2). The pixels help the system locate evidence, not answer questions: System 2 lifts view macro-accuracy to 0.68, against 0.38 for System 1 and 0.14 for an always-CAM_FRONT baseline. Our best run reaches 0.78 joint accu racy, 0.83 view accuracy and 0.65 view macro accuracy on the official test split. Answer accu racy already reaches 0.95 with no image at all: option phrasing alone largely solves the answer head, leaving nearly all the measurable signal in the view head.
%U https://aclanthology.org/2026.groundlm-1.90/
%P 1214-1220
Markdown (Informal)
[Ensar at GroundLM 2026 Shared Tasks: Symbolic Scene Serialisation for View-Level Evidence Identification](https://aclanthology.org/2026.groundlm-1.90/) (Gaafar & Saqr, GroundLM 2026)
ACL