Enregistré dans:
Détails bibliographiques
Auteurs principaux: He, Hulingxiao, Tan, Zhi, Peng, Yuxin
Format: Preprint
Publié: 2026
Sujets:
Accès en ligne:https://arxiv.org/abs/2603.00431
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866917355203854336
author He, Hulingxiao
Tan, Zhi
Peng, Yuxin
author_facet He, Hulingxiao
Tan, Zhi
Peng, Yuxin
contents A high-performing, general-purpose visual understanding model should map visual inputs to a taxonomic tree of labels, identify novel categories beyond the training set for which few or no publicly available images exist. Large Multimodal Models (LMMs) have achieved remarkable progress in fine-grained visual recognition (FGVR) for known categories. However, they remain limited in hierarchical visual recognition (HVR) that aims at predicting consistent label paths from coarse to fine categories, especially for novel categories. To tackle these challenges, we propose Taxonomy-Aware Representation Alignment (TARA), a simple yet effective strategy to inject taxonomic knowledge into LMMs. TARA leverages representations from biology foundation models (BFMs) that encode rich biological relationships through hierarchical contrastive learning. By aligning the intermediate representations of visual features with those of BFMs, LMMs are encouraged to extract discriminative visual cues well structured in the taxonomy tree. Additionally, we align the representations of the first answer token with the ground-truth label, flexibly bridging the gap between contextualized visual features and categories of varying granularity according to user intent. Experiments demonstrate that TARA consistently enhances LMMs' hierarchical consistency and leaf node accuracy, enabling reliable recognition of both known and novel categories within complex biological taxonomies. Code is available at https://github.com/PKU-ICST-MIPL/TARA_CVPR2026.
format Preprint
id arxiv_https___arxiv_org_abs_2603_00431
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Taxonomy-Aware Representation Alignment for Hierarchical Visual Recognition with Large Multimodal Models
He, Hulingxiao
Tan, Zhi
Peng, Yuxin
Computer Vision and Pattern Recognition
Artificial Intelligence
A high-performing, general-purpose visual understanding model should map visual inputs to a taxonomic tree of labels, identify novel categories beyond the training set for which few or no publicly available images exist. Large Multimodal Models (LMMs) have achieved remarkable progress in fine-grained visual recognition (FGVR) for known categories. However, they remain limited in hierarchical visual recognition (HVR) that aims at predicting consistent label paths from coarse to fine categories, especially for novel categories. To tackle these challenges, we propose Taxonomy-Aware Representation Alignment (TARA), a simple yet effective strategy to inject taxonomic knowledge into LMMs. TARA leverages representations from biology foundation models (BFMs) that encode rich biological relationships through hierarchical contrastive learning. By aligning the intermediate representations of visual features with those of BFMs, LMMs are encouraged to extract discriminative visual cues well structured in the taxonomy tree. Additionally, we align the representations of the first answer token with the ground-truth label, flexibly bridging the gap between contextualized visual features and categories of varying granularity according to user intent. Experiments demonstrate that TARA consistently enhances LMMs' hierarchical consistency and leaf node accuracy, enabling reliable recognition of both known and novel categories within complex biological taxonomies. Code is available at https://github.com/PKU-ICST-MIPL/TARA_CVPR2026.
title Taxonomy-Aware Representation Alignment for Hierarchical Visual Recognition with Large Multimodal Models
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2603.00431