Learning to Reduce Dual-Level Discrepancy for Infrared-Visible Person Re-Identification

Wang, Zhixiang; Wang, Zheng; Zheng, Yinqiang; Chuang, Yung-Yu; Satoh, Shin'ich

doi:10.1109/cvpr.2019.00071

Cited by 344 publications

(242 citation statements)

References 19 publications

Supporting

Mentioning

220

Contrasting

Order By: Relevance

“…It mainly has two advantages: 1) end-to-end feature learning directly from the data without extra metric learning steps, 2) it simultaneously handles the cross-modality and intra-modality variations by the dual-constrained top-ranking loss to ensure the discriminability of the learnt person features. Learning to Reduce Dual-level Discrepancy for Infrared-Visible Person Re-identification [26]. Wang et al proposed a novel dual-level discrepancy reduction learning (D 2 RL) scheme to separately handle the two discrepancies: modality discrepancy and appearance discrepancy.…”

Section: Cross-modality Person Re-identification With Generativementioning

confidence: 99%

Enhancing the discriminative feature learning for visible-thermal cross-modality person re-identification

et al. 2020

View full text Add to dashboard Cite

Existing person re-identification has achieved great progress in the visible domain, capturing all the person images with visible cameras. However, in a 24-hour intelligent surveillance system, the visible cameras may be noneffective at night. In this situation, thermal cameras are the best supplemental components, which capture images without depending on visible light. Therefore, in this paper, we investigate the visible-thermal cross-modality person re-identification (VT Re-ID) problem. In VT Re-ID, there are two knotty problems should be well handled, cross-modality discrepancy and intra-modality variations. To address these two issues, we propose focusing on enhancing the discriminative feature learning (EDFL) with two extreme simple means from two core aspects, (1) skip-connection for mid-level features incorporation to improve the person features with more discriminability and robustness, and (2) dual-modality triplet loss to guide the training procedures by simultaneously considering the cross-modality discrepancy and intra-modality variations. Additionally, the two-stream CNN structure is adopted to learn the multi-modality sharable person features. The experimental results on two datasets show that our proposed EDFL approach distinctly outperforms state-of-the-art methods by large margins, demonstrating the effectiveness of our EDFL to enhance the discriminative feature learning for VT Re-ID.

show abstract

Section: Cross-modality Person Re-identification With Generativementioning

confidence: 99%

Enhancing the discriminative feature learning for visible-thermal cross-modality person re-identification

et al. 2020

View full text Add to dashboard Cite

show abstract

“…Wang et al [43] performed a comprehensive survey of heterogeneous person re-identification. Specifically, several types of cross-modality person ReID have been studied, including Image-to-Text cross-modality retrieval [24], Photo-to-Sketch cross-modality retrieval [34], and popular Infrared-to-Visible cross-modality retrieval [8,15,21,41,44,50,54,55,58]. Li et al [24] proposed that searching a person with free-form natural language descriptions can be widely applied in video surveillance and build a dataset for image-text cross-modality retrieval.…”

Section: Cross-modality Retrievalmentioning

confidence: 99%

“…features in a shared space. Wang et al [44] proposed D 2 RL that consists of two GANs to conduct mutual translations of infrared and visible images. On the other hand, AlignGAN [41] accomplished pixel and feature alignment among the visible images, infrared images, and generated fake infrared images within a unified GAN framework.…”

Section: Cross-modality Retrievalmentioning

confidence: 99%

See 1 more Smart Citation

Co-Attentive Lifting for Infrared-Visible Person Re-Identification

Wei

Hong

et al. 2020

Proceedings of the 28th ACM International Conference on Multimedia

View full text Add to dashboard Cite

Infrared-visible cross-modality person re-identification (IV-ReID) has attracted much attention with the popularity of dual-mode video surveillance systems, where the RGB mode works in the daytime and automatically switches to the infrared mode at night. Despite its significant application value, IV-ReID remains a difficult problem mainly due to two great challenges. First, it is difficult to identify persons in the infrared image, which lacks color and texture clues. Second, there is a significant gap between the infrared and visible modalities where appearances of the same person vary considerably. This paper proposes a novel attention-based approach to handle the two difficulties in a unified framework. 1) We propose an attention lifting mechanism to learn discriminative features in each modality. 2) We propose a co-attentive learning mechanism to bridge the gap between the two modalities. Our method only makes slight modifications of a given backbone network and requires small computation overhead while improving the performance significantly. We conduct extensive experiments to demonstrate the superiority of our proposed method. CCS CONCEPTS • Computing methodologies → Visual content-based indexing and retrieval; Appearance and texture representations; • Information systems → Nearest-neighbor search.

show abstract

“…Some studies have used deep learning for cross-modality face recognition between RGB, NIR, and thermal. These provide an elegant solution to the complex problem by utilizing multiple networks [5][6][7][8]. However, there are very few deep-learning studies that focus solely on NIR face recognition.…”

Section: Introductionmentioning

confidence: 99%

NIR Reflection Augmentation for DeepLearning-Based NIR Face Recognition

Kim

2019

Symmetry

View full text Add to dashboard Cite

Face recognition using a near-infrared (NIR) sensor is widely applied to practical applications such as mobile unlocking or access control. However, unlike RGB sensors, few deep learning approaches have studied NIR face recognition. We conducted comparative experiments for the application of deep learning to NIR face recognition. To accomplish this, we gathered five public databases and trained two deep learning architectures. In our experiments, we found that simple architecture could have a competitive performance on the NIR face databases that are mostly composed of frontal face images. Furthermore, we propose a data augmentation method to train the architectures to improve recognition of users who wear glasses. With this augmented training set, the recognition rate for users who wear glasses increased by up to 16%. This result implies that the recognition of those who wear glasses can be overcome using this simple method without constructing an additional training set. Furthermore, the model that uses augmented data has symmetry with those trained with real glasses-wearing data regarding the recognition of people who wear glasses.

show abstract

Learning to Reduce Dual-Level Discrepancy for Infrared-Visible Person Re-Identification

Cited by 344 publications

References 19 publications

Enhancing the discriminative feature learning for visible-thermal cross-modality person re-identification

Enhancing the discriminative feature learning for visible-thermal cross-modality person re-identification

Co-Attentive Lifting for Infrared-Visible Person Re-Identification

NIR Reflection Augmentation for DeepLearning-Based NIR Face Recognition

Contact Info

Product

Resources

About