Detector ensembles for face recognition in video surveillance

Pagano, C.; Granger, Éric; Sabourin, Robert; Gorodnichy, Dmitry O.

doi:10.1109/ijcnn.2012.6252659

Cited by 33 publications

(30 citation statements)

References 22 publications

(33 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…During operation, probe ROI patterns captured from videos are matched against facial models within the gallery. In this context, FR is typically achieved using a system comprised of modules for segmentation, tracking, feature extraction, classification, and spatiotemporal fusion (see Figure 1) [42]. In these systems, each video camera in a network captures the environment in its field of view that may contain individuals during operational phase.…”

Section: Systems For Still-to-video Frmentioning

confidence: 99%

Robust watch-list screening using dynamic ensembles of SVMs based on multiple face representations

Bashbaghi

Granger

Sabourin

et al. 2017

Machine Vision and Applications

Self Cite

View full text Add to dashboard Cite

Abstract. Still-to-video face recognition (FR) is an important function in video surveillance, where faces captured over a network of video cameras are matched against reference stills of target individuals. Screening faces against a watch-list is a challenging video surveillance application because the appearance of faces vary due to changing capture conditions and operational domains. The facial models used for matching may not be representative of faces captured with video cameras because they are typically designed a priori with only one reference still. In this paper, a multi-classifier framework is proposed for robust still-to-video FR based on multiple and diverse face representations of a single reference face still. During enrollment of a target individual, the single reference face still is modeled using an ensemble of SVM classifiers based on different patches and face descriptors. Multiple feature extraction techniques are applied to patches isolated in the reference still to generate a diverse SVM pool that provides robustness to common nuisance factors (e.g., variations in illumination and pose). The estimation of discriminant feature subsets, classifier parameters, decision thresholds, and ensemble fusion functions is achieved using the high-quality reference still and a large number of faces captured in lower quality video of non-target individuals in the scene. During operations, the most competent subset of SVMs are dynamically selected according to capture conditions. Finally, a head-face tracker gradually regroups faces captured from different people appearing in a scene, while each individual-specific ensemble performs face matching. The accumulation of matching scores per face track leads to a robust spatio-temporal FR when accumulated ensemble scores surpass a detection threshold. Experimental results obtained with the Chokepoint and COX-S2V datasets show a significant improvement in performance w.r.t. reference systems, especially when individual-specific ensembles (1) are designed using exemplar-SVMs rather than one-class SVMs, and (2) exploit score-level fusion of local SVMs (trained using features extracted from each patch), rather than using either decision-level or feature-level fusion with a global SVM (trained by concatenating features extracted from patches).

show abstract

Section: Systems For Still-to-video Frmentioning

confidence: 99%

Robust watch-list screening using dynamic ensembles of SVMs based on multiple face representations

Bashbaghi

Granger

Sabourin

et al. 2017

Machine Vision and Applications

Self Cite

View full text Add to dashboard Cite

show abstract

“…EoDs are co-jointly trained using a dynamic particle swarm optimization (DPSO) based training strategy, generating a diversified pool of ARTMAP neural networks. Trained detectors are selected and combined using boolean combination (BC) [17].…”

Section: Adaptive Face Recognition In Videomentioning

confidence: 99%

An Individual-Specific Strategy for Management of Reference Data in Adaptive Ensembles for Person Re-Identification

De-la-Torre¹,

Granger²,

Sabourin³

et al. 2013

5th International Conference on Imaging for Crime Detection and Prevention (ICDP 2013)

View full text Add to dashboard Cite

In video surveillance, person re-identification refers to recognizing individuals of interest from faces captured across a network of video cameras. Face recognition in such applications is challenging because faces are captured with limited spatial and temporal constraints. In addition, facial models for recognition are commonly designed using limited reference samples from faces captured under specific conditions. Given new reference samples, updating facial models may allow maintaining a high level of performance over time. Although adaptive ensembles have been successfully applied to robust modeling of an individual's face, reference data samples must be stored for validation. In this paper, a memory management strategy based on Kullback-Leiber (KL) divergence is proposed to rank and select the most relevant validation samples over time in adaptive individual-specific ensembles. When new reference data becomes available for an individual, updates to the corresponding ensembles are validated using a mixture of new and previously-stored samples. Only the samples with the highest KL divergence are preserved in memory for future adaptations. The strategy is compared with reference classifiers using videos from the FIA data set. Simulation results show that the proposed strategy tends to select samples of statistically different subjects (so-called "wolfs") for validation, thereby reducing the number of samples per individual by up to 80%, yet maintaining a high level of performance.

show abstract

“…Systems for still-to-video FR applied to VS are typically modeled in terms of independent detection problems [5], each one implemented using template matching or a one-or two-class classifier per individual followed by thresholding. These individual-specific detectors are designed with reference ROI patterns from target, and possibly non-target individuals (from the cohort or universal background).…”

Section: Introductionmentioning

confidence: 99%

“…These individual-specific detectors are designed with reference ROI patterns from target, and possibly non-target individuals (from the cohort or universal background). The advantages of such modular architectures include the ease with which face models may be added, updated and removed from the systems, and the possibility of specializing feature subsets and decision thresholds to each specific individual [5].…”

Section: Introductionmentioning

confidence: 99%

Adaptive appearance model tracking for still-to-video face recognition

Dewan

Granger

Marcialis

et al. 2016

Pattern Recognition

View full text Add to dashboard Cite

Systems for still-to-video face recognition (FR) seek to detect the presence of target individuals based on reference facial still images or mug-shots. These systems encounter several challenges in video surveillance applications due to variations in capture conditions (e.g., pose, scale, illumination, blur and expression) and to camera inter-operability. Beyond these issues, few reference stills are available during enrollment to design representative facial models of target individuals. Systems for still-to-video FR must therefore rely on adaptation, multiple face representation, or synthetic generation of reference stills to enhance the intra-class variability of face models. Moreover, many FR systems only match high quality faces captured in video, which further reduces the probability of detecting target individuals. Instead of matching faces captured through segmentation to reference stills, this paper exploits Adaptive Appearance Model Tracking (AAMT) to gradually learn a track-face-model for each individual appearing in the scene. The Sequential Karhunen-Loeve technique is used for online learning of these track-face-models within a particle filter-based face tracker. Meanwhile, these models are matched over successive frames against the reference still images of each target individual enrolled to the system, and then matching scores are accumulated over several frames for robust spatiotemporal recognition. A target individual is recognized if scores accumulated for a track-face-model over a fixed time surpass some decision threshold. The main advantage of AAMT over traditional still-to-video FR systems is the greater diversity of facial representation that may be captured during operations, and this can lead to better discrimination for spatiotemporal recognition. Compared to state-of-the-art adaptive biometric systems, the proposed method selects facial captures to update an individual's face model more reliably because it relies on information from tracking. Simulation results obtained with the Chokepoint video dataset indicate that the proposed method provides a significantly higher level of performance compared state-of-the-art systems when a single reference still per individual is available for matching. This higher level of performance is achieved when the diverse facial appearances that are captured in video through AAMT correspond to that of reference stills.

show abstract

Detector ensembles for face recognition in video surveillance

Cited by 33 publications

References 22 publications

Robust watch-list screening using dynamic ensembles of SVMs based on multiple face representations

Robust watch-list screening using dynamic ensembles of SVMs based on multiple face representations

An Individual-Specific Strategy for Management of Reference Data in Adaptive Ensembles for Person Re-Identification

Adaptive appearance model tracking for still-to-video face recognition

Contact Info

Product

Resources

About