TRIM-ReID: Duplication-Aware Token Reduction and Modality-Aligned Interaction for Multi-Modal Object Re-Identification

Wanke Xia; Ruiding Zhu; Xingguo Xu; Zhengbo Zhang; Dongxia Liu; Yuan Jin; Taojie Zhu; Yiting Zhao

cs.CVOct 3, 2026

TRIM-ReID: Duplication-Aware Token Reduction and Modality-Aligned Interaction for Multi-Modal Object Re-Identification

Authors: Wanke Xia, Ruiding Zhu, Xingguo Xu, Zhengbo Zhang, Dongxia Liu, Yuan Jin, Taojie Zhu, Yiting Zhao,

Organizations: Tsinghua University · Anhui University · Dalian University of Technology · CASIA ·

Abstract

Multi-modal object re-identification exploits complementary RGB, near-infrared (NIR), and thermal-infrared (TIR) observations to retrieve target objects. However, existing methods commonly employ visual encoders optimized for global image-text alignment and select tokens using learned importance scores. Such designs fail to preserve fine-grained identity cues or explicitly account for token redundancy, resulting in underrepresented local evidence and duplicated tokens that lead to noisy and costly cross-modal interaction. To address this gap, we propose TRIM-ReID, a compact framework that unifies dense feature extraction, intra-modal token reduction, and inter-modal aligned interaction. Specifically, semantically rich and spatially coherent patch features are extracted by Dense Identity Representation (DIR), which leverages DINOv3 to preserve fine-grained identity information. We then introduce Token Diversity Mining (TDM) to identify complementary local evidence and construct compact modality-specific token sets by suppressing repetitive patches while preserving informative diversity. Retained tokens are subsequently fused by Modal Relational Interaction (MRI) to enable effective information exchange across modalities, while a triangular alignment loss explicitly regularizes their joint relationships to maintain cross-modal semantic consistency under independent token selection. Extensive experiments on RGBNT201, RGBNT100, and MSVR310 demonstrate that TRIM-ReID achieves state-of-the-art performance.

Figures & tables

Figure 1: Overall framework of TRIM-ReID. It comprises three core components: DIR, TDM and MRI.

Figure 2: Visualization of attention maps and token reduction.

Methods	Venue	RGBNT201				RGBNT100		MSVR310
Methods	Venue	mAP	R-1	R-5	R-10	mAP	R-1	mAP	R-1
PFNet [ 16 ]	AAAI’21	38.5	38.9	52.0	58.4	68.1	94.1	23.5	37.4
TOP-ReID [ 8 ]	AAAI’24	72.3	76.6	84.7	89.4	81.2	96.4	35.9	44.6
EDITOR [ 13 ]	CVPR’24	66.5	68.3	81.1	88.2	82.1	96.4	39.0	49.3
LRMM [ 11 ]	ESWA’25	52.3	53.4	64.6	73.2	78.6	96.7	36.7	49.7
IDEA [ 9 ]	CVPR’25	80.2	82.1	90.0	93.3	87.2	96.5	47.0	62.4

Table 1: Performance comparison of various methods on the datasets of RGBNT201, RGBNT100, and MSVR310. The best and second results are in bold and underlined , respectively. Higher values indicate better performance for all metrics.

Variants	Modules			Metrics
Variants	DIR	TDM	MRI	mAP	R-1
A	$\checkmark$	$\times$	$\times$	76.4	79.1
B	$\checkmark$	$\checkmark$	$\times$	80.6	84.6
C	$\checkmark$	$\checkmark$	$\checkmark$	85.2	88.0

Table 2: Ablation on the proposed modules of TRIM-ReID.

Encoders	mAP	R-1	R-5	R-10
CLIP [ 4 ]	74.8	80.5	87.2	90.8
DINOv3 [ 5 ]	85.2	88.0	92.7	95.5

Table 3: Comparison of different visual encoders.

Number Setting	mAP	R-1	R-5	R-10
$N_{r}=32$	81.5	84.8	89.1	91.9
$N_{r}=48$	82.6	89.0	93.0	95.0
$N_{r}=64$	80.3	83.4	89.6	92.7
$N_{r}=80$	85.2	88.0	92.7	95.5
$N_{r}=96$	80.7	83.9	91.3	94.1
$N_{r}=112$	83.3	87.0	91.5	94.0

Table 4: Ablation on the number of retained tokens in TDM.

Number Setting	mAP	R-1	R-5	R-10
$N_{p}=2$	82.1	82.8	90.7	94.4
$N_{p}=4$	85.2	88.0	92.7	95.5
$N_{p}=8$	83.7	86.4	92.7	93.9

Table 5: Ablation on the number of pivot tokens in TDM.

Multimodal Fusion

Multimodal fusion combines information from different data types to improve prediction, perception, and decision-making.

See all 771 papers

Feb 28, 2026

STMI: Segmentation-Guided Token Modulation with Cross-Modal Hypergraph Interaction for Multi-Modal Object Re-Identification

Multi-modal object Re-Identification (ReID) aims to exploit complementary information from different modalities to retrieve specific objects. However, existing methods often rely on hard token filtering or simple fusion strategies, which can lead to the loss of discriminative cues and increased background interference. To address these challenges, we propose STMI, a novel multi-modal learning framework consisting of three key components: (1) Segmentation-Guided Feature Modulation (SFM) module leverages SAM-generated masks to enhance foreground representations and suppress background noise through learnable attention modulation; (2) Semantic Token Reallocation (STR) module employs learnable query tokens and an adaptive reallocation mechanism to extract compact and informative representations without discarding any tokens; (3) Cross-Modal Hypergraph Interaction (CHI) module constructs a unified hypergraph across modalities to capture high-order semantic relationships. Extensive experiments on public benchmarks (i.e., RGBNT201, RGBNT100, and MSVR310) demonstrate the effectiveness and robustness of our proposed STMI framework in multi-modal ReID scenarios.

Xingguo Xu, Zhanyu Liu, Weixiang Zhou +5

School of Mathematical Sciences, Dalian University of Technology, China · College of Computer Science and Technology, Zhejiang University, China · School of Future Technology, Dalian University of Technology, China +1

Cross-Modal Representation Learning

Ask PDF

Jul 26, 2026

Multi-Modal Object Re-Identification with Prompt-S6 and Semantic-Aware Knowledge Guidance

Multi-modal object Re-Identification (ReID) aims to retrieve specific objects by integrating complementary information from multiple modalities. However, existing multi-modal ReID methods do not effectively address background interference suppression or achieve tri-modal alignment, instead focusing on pairwise feature fusion. Moreover, many current aggregation approaches suffer from high computational complexity. To address these limitations, we propose PRISM, a novel multi-modal ReID framework built upon Prompt-S6 (PS6) and semantic-aware knowledge guidance. PS6 maintains the linear complexity and strong sequence modeling capability of Mamba while enabling efficient cross-modal interaction. Leveraging these advantages, we design two key components: Semantic-Driven Token Pruning (SDTP) and Progressive Fusion Network (PFN). Parsing semantic priors from the segmentation foundation models, the SDTP then leverages these priors and applies dynamic token pruning to suppress background noise and refine feature representations. The PFN progressively aggregates multi-modal features to achieve tri-modal alignment and fully exploit modality complementarity. With the proposed modules, PRISM generates more robust multi-modal representations under complex scenarios. Extensive experiments on four multi-modal object ReID benchmarks demonstrate the effectiveness and efficiency of our approach. The source code is available at https://github.com/zw-absin/PRISM.

Weixiang Zhou, Jiabei Zuo, Yuhao Wang +3

School of Mathematical Sciences and the Key Laboratory for Computational Mathematics and Data Intelligence of Liaoning Province, Dalian University of Technology, Dalian, 116024, China · School of Computer Science and Technology, Dalian University of Technology, Dalian, 116024, China · School of Future Technology, School of Artificial Intelligence, Dalian University of Technology, Dalian, 116024, China +1

Cross-Modal Alignment Multimodal Fusion

Ask PDF

Jul 31, 2026

Multi-Modal Object Re-Identification with Dual Semantic Guidance and Global-Local Mutual Modulation

Multi-modal object Re-Identification (ReID) aims to retrieve target instances by leveraging complementary information across modalities. However, existing methods suffer from two challenges. First, they often fail to exploit well-aligned and reliable semantic priors, making them vulnerable to background clutter and cross-modal misalignment. On the other hand, they typically rely on holistic feature modeling, overlooking the synergy between global and local representations. To overcome these limitations, we propose a robust multi-modal ReID framework with dual semantic guidance and global-local mutual modulation, which mainly consists of three key components, namely the Text-Semantic Injector (TSI), the Masked Global-Local Modulator (MGLM), and the Hierarchical MoE Fusion (HMF). The TSI enhances semantic awareness by integrating clean and coherent textual features into visual tokens. The MGLM enables part-aware cross-modal interaction through joint guidance from soft masks and global context, improving fine-grained feature alignment. Finally, the HMF adaptively aggregates multi-spectral features under local semantic supervision, yielding discriminative and robust representations. Extensive experiments on three multi-modal ReID benchmarks demonstrate the effectiveness of the proposed method. The code will be made publicly available at https://github.com/zw-absin/DSGM upon acceptance.

Weixiang Zhou, Xingguo Xu, Yuhao Wang +4

School of Mathematical Sciences and the Key Laboratory for Computational Mathematics and Data Intelligence of Liaoning Province, Dalian University of Technology, Dalian, 116024, China · School of Future Technology, School of Artificial Intelligence, Dalian University of Technology, Dalian, 116024, China · Department of Radiology and Biomedical Imaging, University of California, San Francisco, 94107, USA +1

Cross-Modal Representation Learning Multimodal Robustness Cross-Modal Alignment Multimodal Fusion

Ask PDF

Abstract

Multimodal Fusion

Explore similar work

STMI: Segmentation-Guided Token Modulation with Cross-Modal Hypergraph Interaction for Multi-Modal Object Re-Identification

Multi-Modal Object Re-Identification with Prompt-S6 and Semantic-Aware Knowledge Guidance

Multi-Modal Object Re-Identification with Dual Semantic Guidance and Global-Local Mutual Modulation