Salvato in:
Dettagli Bibliografici
Autori principali: Li, Jialin, Wu, Shuqi, Wang, Ning
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:https://arxiv.org/abs/2508.11218
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866913993212297216
author Li, Jialin
Wu, Shuqi
Wang, Ning
author_facet Li, Jialin
Wu, Shuqi
Wang, Ning
contents Re-Identification (ReID) is a critical technology in intelligent perception systems, especially within autonomous driving, where onboard cameras must identify pedestrians across views and time in real-time to support safe navigation and trajectory prediction. However, the presence of uncertain or missing input modalities--such as RGB, infrared, sketches, or textual descriptions--poses significant challenges to conventional ReID approaches. While large-scale pre-trained models offer strong multimodal semantic modeling capabilities, their computational overhead limits practical deployment in resource-constrained environments. To address these challenges, we propose a lightweight Uncertainty Modal Modeling (UMM) framework, which integrates a multimodal token mapper, synthetic modality augmentation strategy, and cross-modal cue interactive learner. Together, these components enable unified feature representation, mitigate the impact of missing modalities, and extract complementary information across different data types. Additionally, UMM leverages CLIP's vision-language alignment ability to fuse multimodal inputs efficiently without extensive finetuning. Experimental results demonstrate that UMM achieves strong robustness, generalization, and computational efficiency under uncertain modality conditions, offering a scalable and practical solution for pedestrian re-identification in autonomous driving scenarios.
format Preprint
id arxiv_https___arxiv_org_abs_2508_11218
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle A CLIP-based Uncertainty Modal Modeling (UMM) Framework for Pedestrian Re-Identification in Autonomous Driving
Li, Jialin
Wu, Shuqi
Wang, Ning
Computer Vision and Pattern Recognition
Machine Learning
Re-Identification (ReID) is a critical technology in intelligent perception systems, especially within autonomous driving, where onboard cameras must identify pedestrians across views and time in real-time to support safe navigation and trajectory prediction. However, the presence of uncertain or missing input modalities--such as RGB, infrared, sketches, or textual descriptions--poses significant challenges to conventional ReID approaches. While large-scale pre-trained models offer strong multimodal semantic modeling capabilities, their computational overhead limits practical deployment in resource-constrained environments. To address these challenges, we propose a lightweight Uncertainty Modal Modeling (UMM) framework, which integrates a multimodal token mapper, synthetic modality augmentation strategy, and cross-modal cue interactive learner. Together, these components enable unified feature representation, mitigate the impact of missing modalities, and extract complementary information across different data types. Additionally, UMM leverages CLIP's vision-language alignment ability to fuse multimodal inputs efficiently without extensive finetuning. Experimental results demonstrate that UMM achieves strong robustness, generalization, and computational efficiency under uncertain modality conditions, offering a scalable and practical solution for pedestrian re-identification in autonomous driving scenarios.
title A CLIP-based Uncertainty Modal Modeling (UMM) Framework for Pedestrian Re-Identification in Autonomous Driving
topic Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2508.11218