What "Not" to Detect: Negation-Aware VLMs via Structured Reasoning and Token Merging

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Kang, Inha, Lim, Youngsun, Lee, Seonho, Choi, Jiho, Choe, Junsuk, Shim, Hyunjung
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866908906263937024
author Kang, Inha
Lim, Youngsun
Lee, Seonho
Choi, Jiho
Choe, Junsuk
Shim, Hyunjung
author_facet Kang, Inha
Lim, Youngsun
Lee, Seonho
Choi, Jiho
Choe, Junsuk
Shim, Hyunjung
contents State-of-the-art vision-language models (VLMs) suffer from a critical failure in understanding negation, often referred to as affirmative bias. This limitation is particularly severe in described object detection (DOD) tasks. To address this, we propose two primary contributions: (1) a new dataset pipeline and (2) a novel, lightweight adaptation recipe. First, we introduce CoVAND, a dataset constructed with a systematic chain-of-thought (CoT) and VQA-based pipeline to generate high-quality, instance-grounded negation data. Second, we propose NegToMe, a novel text token merging module that directly tackles the architectural cause of affirmative bias. NegToMe fundamentally addresses the structural loss of negation cues in tokenization, grouping them with attributes into coherent semantic phrases. It maintains correct polarity at the input level, enabling robust negation understanding even with limited data. For instance, to prevent a model from treating the fragmented tokens "not" and "girl" as simply "girl", NegToMe binds them into a single token whose meaning is correctly distinguished from that of "girl" alone. This module is integrated with a parameter-efficient and strategic LoRA fine-tuning approach. Our method significantly improves performance on challenging negation benchmarks with a lowered false positive rate, boosting NMS-AP by up to +10.8 points on OVDEval and demonstrating generalization to SoTA VLMs. This work marks a crucial step forward in addressing negation understanding for real-world detection applications.
format Preprint
id arxiv_https___arxiv_org_abs_2510_13232
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle What "Not" to Detect: Negation-Aware VLMs via Structured Reasoning and Token Merging
Kang, Inha
Lim, Youngsun
Lee, Seonho
Choi, Jiho
Choe, Junsuk
Shim, Hyunjung
Computer Vision and Pattern Recognition
Artificial Intelligence
State-of-the-art vision-language models (VLMs) suffer from a critical failure in understanding negation, often referred to as affirmative bias. This limitation is particularly severe in described object detection (DOD) tasks. To address this, we propose two primary contributions: (1) a new dataset pipeline and (2) a novel, lightweight adaptation recipe. First, we introduce CoVAND, a dataset constructed with a systematic chain-of-thought (CoT) and VQA-based pipeline to generate high-quality, instance-grounded negation data. Second, we propose NegToMe, a novel text token merging module that directly tackles the architectural cause of affirmative bias. NegToMe fundamentally addresses the structural loss of negation cues in tokenization, grouping them with attributes into coherent semantic phrases. It maintains correct polarity at the input level, enabling robust negation understanding even with limited data. For instance, to prevent a model from treating the fragmented tokens "not" and "girl" as simply "girl", NegToMe binds them into a single token whose meaning is correctly distinguished from that of "girl" alone. This module is integrated with a parameter-efficient and strategic LoRA fine-tuning approach. Our method significantly improves performance on challenging negation benchmarks with a lowered false positive rate, boosting NMS-AP by up to +10.8 points on OVDEval and demonstrating generalization to SoTA VLMs. This work marks a crucial step forward in addressing negation understanding for real-world detection applications.
title What "Not" to Detect: Negation-Aware VLMs via Structured Reasoning and Token Merging
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2510.13232