@inproceedings{an-etal-2026-talk,
title = "Talk in Pieces, See in Whole: Disentangled and Hierarchical Representation Learning in Language-based Object Detection",
author = "An, Sojung and
Park, Kwanyong and
Lee, Yong Jae and
Kim, Donghyun",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-main.1793/",
pages = "40850--40874",
ISBN = "979-8-89176-459-0",
abstract = "Vision-language models (VLMs) have advanced multimodal perception, demonstrated by open-vocabulary object detection with simple language queries. State-of-the-art VLMs still struggle to handle complex queries involving descriptive attributes and relational clauses. To address this problem, we propose restructuring linguistic representations according to the hierarchical relations within sentences for language-based object detection. A key insight is the necessity of disentangling textual tokens into core components{---}objects, attributes, and relations{---}and subsequently aggregating them into hierarchically structured sentence-level representations. Building on this principle, we introduce the TaSe (Talk in Pieces, See in Whole) framework with three main contributions: (1) a hierarchical synthetic captioning dataset spanning three tiers from category names to descriptive sentences; (2) the three-component disentanglement module guided by a novel disentanglement loss function, transforms text embeddings into subspace compositions; and (3) aggregation disentangled components into hierarchically structured embeddings with the guide of proposed hierarchical objectives. Experimental results under the OmniLabel benchmark show a 24{\%} performance improvement, demonstrating the importance of linguistic compositionality."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="an-etal-2026-talk">
<titleInfo>
<title>Talk in Pieces, See in Whole: Disentangled and Hierarchical Representation Learning in Language-based Object Detection</title>
</titleInfo>
<name type="personal">
<namePart type="given">Sojung</namePart>
<namePart type="family">An</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kwanyong</namePart>
<namePart type="family">Park</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yong</namePart>
<namePart type="given">Jae</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Donghyun</namePart>
<namePart type="family">Kim</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 2026 Conference on Empirical Methods in Natural Language Processing</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">979-8-89176-459-0</identifier>
</relatedItem>
<abstract>Vision-language models (VLMs) have advanced multimodal perception, demonstrated by open-vocabulary object detection with simple language queries. State-of-the-art VLMs still struggle to handle complex queries involving descriptive attributes and relational clauses. To address this problem, we propose restructuring linguistic representations according to the hierarchical relations within sentences for language-based object detection. A key insight is the necessity of disentangling textual tokens into core components—objects, attributes, and relations—and subsequently aggregating them into hierarchically structured sentence-level representations. Building on this principle, we introduce the TaSe (Talk in Pieces, See in Whole) framework with three main contributions: (1) a hierarchical synthetic captioning dataset spanning three tiers from category names to descriptive sentences; (2) the three-component disentanglement module guided by a novel disentanglement loss function, transforms text embeddings into subspace compositions; and (3) aggregation disentangled components into hierarchically structured embeddings with the guide of proposed hierarchical objectives. Experimental results under the OmniLabel benchmark show a 24% performance improvement, demonstrating the importance of linguistic compositionality.</abstract>
<identifier type="citekey">an-etal-2026-talk</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.1793/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>40850</start>
<end>40874</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Talk in Pieces, See in Whole: Disentangled and Hierarchical Representation Learning in Language-based Object Detection
%A An, Sojung
%A Park, Kwanyong
%A Lee, Yong Jae
%A Kim, Donghyun
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F an-etal-2026-talk
%X Vision-language models (VLMs) have advanced multimodal perception, demonstrated by open-vocabulary object detection with simple language queries. State-of-the-art VLMs still struggle to handle complex queries involving descriptive attributes and relational clauses. To address this problem, we propose restructuring linguistic representations according to the hierarchical relations within sentences for language-based object detection. A key insight is the necessity of disentangling textual tokens into core components—objects, attributes, and relations—and subsequently aggregating them into hierarchically structured sentence-level representations. Building on this principle, we introduce the TaSe (Talk in Pieces, See in Whole) framework with three main contributions: (1) a hierarchical synthetic captioning dataset spanning three tiers from category names to descriptive sentences; (2) the three-component disentanglement module guided by a novel disentanglement loss function, transforms text embeddings into subspace compositions; and (3) aggregation disentangled components into hierarchically structured embeddings with the guide of proposed hierarchical objectives. Experimental results under the OmniLabel benchmark show a 24% performance improvement, demonstrating the importance of linguistic compositionality.
%U https://aclanthology.org/2026.emnlp-main.1793/
%P 40850-40874
Markdown (Informal)
[Talk in Pieces, See in Whole: Disentangled and Hierarchical Representation Learning in Language-based Object Detection](https://aclanthology.org/2026.emnlp-main.1793/) (An et al., EMNLP 2026)
ACL