A Joint Deep Boltzmann Machine (jDBM) Model for Person Identification Using Mobile Phone Data

Alam, Mohammad Rafiqul; Bennamoun, Mohammed; Togneri, Roberto; Sohel, Ferdous

doi:10.1109/tmm.2016.2615524

Cited by 25 publications

(20 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Person ReID has been attracting a lot of research efforts in the community. Traditionally, it has been addressed to deal with from two aspects: the feature representation and metric learning problem, while current state-of-the-art ReID methods are all based on deep learning techniques [26], [27], [28], [29], [30], [31], [32], [9], which jointly optimizes the two phases together. Zhou et al [7] proposed a set to set distance to learn discriminative and stable feature representations and to effectively find out the matched target to the probe object among various candidates.…”

Section: B Person Reid Methodsmentioning

confidence: 99%

Person Re-Identification in Aerial Imagery

Zhang

Yang

Wei

et al. 2021

IEEE Trans. Multimedia

View full text Add to dashboard Cite

Nowadays, with the rapid development of consumer Unmanned Aerial Vehicles (UAVs), visual surveillance by utilizing the UAV platform has been very attractive. Most of the research works for UAV captured visual data are mainly focused on the tasks of object detection and tracking. However, limited attention has been paid to the task of person Re-identification (ReID) which has been widely studied in ordinary surveillance cameras with fixed emplacements.In this paper, to facilitate the research of person ReID in aerial imagery, we collect a large scale airborne person ReID dataset named as Person ReID for Aerial Imagery (PRAI-1581), which consists of 39,461 images of 1581 person identities. The images of the dataset are captured by two DJI consumer UAVs flying at an altitude ranging from 20 to 60 meters above the ground, which covers most of the real UAV surveillance scenarios. In addition, we propose to utilize subspace pooling with SVD of convolution feature maps to represent the input person images. The proposed method can learn a discriminative and compact feature descriptor for ReID in aerial imagery and can be trained via an end-to-end fashion efficiently. We conduct extensive experiments on our dataset and the experimental results demonstrate that the proposed method achieves state-of-the-art performance.

show abstract

Section: B Person Reid Methodsmentioning

confidence: 99%

Person Re-Identification in Aerial Imagery

Zhang

Yang

Wei

et al. 2021

IEEE Trans. Multimedia

View full text Add to dashboard Cite

show abstract

“…In the field of visual-audio dual-modality biometrics, the general pipelines follow three stages including raw feature extraction, feature fusion for joint representation, and identity decision. Modality fusion [5] can be categorized as feature-level (early) [6], [7], classifier-level (intermediate) [8], [9], [10], [11] or score/decision-level (late) [12], [13], [14] fusion. Existing visual-audio biometrics algorithms are listed in Table I.…”

Section: A Related Workmentioning

confidence: 99%

“…Some joint visual-audio representations concatenate intensity-level visual features with audio features using a summation or maximum rule, while others train a deep fusion model, such as the cross-modal prediction model [10] or the joint Boltzmann machine model [11]. Quantitatively, some algorithms achieve nearly perfect results (higher than 95%) by using more audio or intensity information (more passwords, more face profiles, integrating teeth modality).…”

Section: A Related Workmentioning

confidence: 99%

Dual-modality Talking-metrics: 3D Visual-Audio Integrated Behaviometric Cues from Speakers

Zhang

Richmond

Fisher

2018

2018 24th International Conference on Pattern Recognition (ICPR)

View full text Add to dashboard Cite

Face-based behaviometrics focus on dynamic biological signatures generated from face behaviors, which are informative and subject-specific for identity recognition. Most existing face behaviometrics rely on 2D visual features and thus are sensitive to pose or intensity variations. This paper presents a dual-modality behaviometrics algorithm (talking-metrics) that integrates 3D video and audio cues from a human face speaking a passphrase. Static and dynamic 3D face features are extracted algorithmically and audio features are transformed through a few learning models. We concatenate the top 18 discriminative 3D visual-audio features to represent the bi-modality and utilize an linear discrimant analysis (LDA) classifier for identity recognition. The experiments were conducted on a new publicly released dataset (S3DFM). Both qualitative feature distributions and quantitative comparison results show the feasibility of the proposed pipeline and the superiority over using each modality independently. A 98.5% cross-validation recognition rate over 60 subjects and 10 trials was achieved. An anti-spoofing test also demonstrates the robustness of the proposed method.

show abstract

“…This personal recognition technology, called biometrics, has been proven to be more effective than conventional technologies due to its reliable identification/verification of people. Moreover, it meets various criteria, namely uniqueness, robustness, universality, acceptability, non‐reproducibility, recovery, and performance …”

Section: Introductionmentioning

confidence: 99%

Double watermarking‐based biometric access control for radio frequency identification card

Haddada

Amara

2019

Int J RF Microw Comput Aided Eng

View full text Add to dashboard Cite

The integration of data privacy and security into radio frequency identification (RFID) technology, particularly into RFID tags, has become one of the most attractive research areas. A crucial challenge in RFID technology research lies in providing an efficient protection for systems against information theft and illegitimate access. This article proposes a secure solution based on an RFID card for physical biometric access-control applications. This is done by integrating two biometric modalities, namely face and fingerprint which are secured via a double watermarking technique. The suggested approach is ensured by two levels of watermarking. At the first level, the wavelet packet decomposition watermarking algorithm is used to insert features from the fingerprint (minutiae) in the face image of an authorized person. At the second level, the same watermarking algorithm is employed to insert the fingerprint watermark in the face features extracted by Gabor filters from the previously watermarked face image (at the first level). The obtained secured watermarked biometric data are then integrated in a 1-kB high frequency proximity RFID card. This combination of both RFID technology and the double watermarking technique provides a biometric control access framework.Compared with the state-of-the-art frameworks, the proposed one ensures a good compromise between a reduced computational complexity and a high level of data security while maintaining a small space of storage and a low cost compared to those of the marketed products. K E Y W O R D Sauthentication framework, biometrics, control access, double watermarking, radio frequency identification, security

show abstract

A Joint Deep Boltzmann Machine (jDBM) Model for Person Identification Using Mobile Phone Data

Cited by 25 publications

References 15 publications

Person Re-Identification in Aerial Imagery

Person Re-Identification in Aerial Imagery

Dual-modality Talking-metrics: 3D Visual-Audio Integrated Behaviometric Cues from Speakers

Double watermarking‐based biometric access control for radio frequency identification card

Contact Info

Product

Resources

About