# ProtoHGF-Net: Prototype HyperGraph Fusion with Intra-modal Calibration for RGBT Object Detection

Source: [ProtoHGF-Net: Prototype HyperGraph Fusion with Intra-modal Calibration for RGBT Object Detection](<https://www.sota2.com/research/paper/6a7d4fa94b236a9bba1a22f0>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-08-14T05:00:04.745Z

Authors: Xiangqi Chen, Xiuling Zhang, Chengzhuan Yang, Li Zhao, Dawei Zhang, Yanchao Wang, Liyuan Chen, Hua Wang, Hao Peng, Zhonglong Zheng

Published: 2026-08-12T03:08:19.000Z

## Abstract

RGB-Thermal (RGBT) object detection enables robust perception in complex scenes by leveraging the complementary strengths of visible textures and thermal cues. However, existing methods mainly rely on dense cross-modal interactions over full-resolution features, which inevitably introduce background interference and hinder the learning of target-relevant representations. In this paper, we propose the Prototype HyperGraph Fusion Network (ProtoHGF-Net), a novel framework that redefines cross-modal fusion as prototype-level semantic interaction rather than the dense cross-modal interaction paradigm. Specifically, we design Prototype HyperGraph Fusion to perform cross-modal interaction in a compact prototype-level semantic space. This design enables more selective fusion among target-relevant prototypes. To support this prototype-level fusion, we propose Teacher-Mask Calibration Distillation, which calibrates modality features before fusion using modality-specific teachers and target-aware masks. This strategy suppresses backgrou- nd-dominant responses and produces more target-focused features. Extensive experiments on DroneVehicle, DVTOD, and FLIR demonstrate that ProtoHGF-Net achieves state-of-the-art performance with 85.9\\% $mAP\_{50}$, 88.2\\% $mAP\_{50}$, and 79.1\\% $mAP\_{50}$, respectively. Our code is available at \\href{https://github.com/ZiMo-Chen/ProtoHGF}{GitHub}.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2608.11595>)

[PDF](<https://arxiv.org/pdf/2608.11595.pdf>)