Report on the FG 2015 Video Person Recognition Evaluation

Beveridge, J. Ross; Zhang, Hao; Draper, Bruce A.; Flynn, Patrick J.; Feng, Zhen-Hua; Huber, Patrik; Kittler, Josef; Huang, Zhiwu; Li, Shaoxin; Li, Yan; Kan, Meina; Wang, Ruiping; Shan, Shiguang; Chen, Xilin; Li, Haoxiang; Hua, Gang; Krizaj, Janez; Ding, Changxing; Tao, Dacheng; Phillips, P. Jonathon

doi:10.1109/fg.2015.7163156

Cited by 41 publications

(27 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, facial symmetry does not exactly hold true for high-resolution images where fine facial textures are clear. For low-resolution face images, e.g., frames of surveillance videos Beveridge et al (2015), this strategy may work well.…”

Section: Summary and Concluding Remarksmentioning

confidence: 99%

A Comprehensive Survey on Pose-Invariant Face Recognition

Ding

Tao

2016

ACM Trans. Intell. Syst. Technol.

260

137

View full text Add to dashboard Cite

The capacity to recognize faces under varied poses is a fundamental human ability that presents a unique challenge for computer vision systems. Compared to frontal face recognition, which has been intensively studied and has gradually matured in the past few decades, pose-invariant face recognition (PIFR) remains a largely unsolved problem. However, PIFR is crucial to realizing the full potential of face recognition for real-world applications, since face recognition is intrinsically a passive biometric technology for recognizing uncooperative subjects. In this paper, we discuss the inherent difficulties in PIFR and present a comprehensive review of established techniques. Existing PIFR methods can be grouped into four categories, i.e., pose-robust feature extraction approaches, multi-view subspace learning approaches, face synthesis approaches, and hybrid approaches. The motivations, strategies, pros/cons, and performance of representative approaches are described and compared. Moreover, promising directions for future research are discussed.

show abstract

Section: Summary and Concluding Remarksmentioning

confidence: 99%

A Comprehensive Survey on Pose-Invariant Face Recognition

Ding

Tao

2016

ACM Trans. Intell. Syst. Technol.

260

137

View full text Add to dashboard Cite

show abstract

“…We construct a CSR by progressively training a set of cascaded weak regressors. To this end, we first initialise the face shape estimate, s n , for each training image input : image I and a trained CSR model Φ = {φ 1 , ..., φ M } output: facial landmarks s 1 initialise the current face shape, s , using the detected face bounding box; 2 for m ← 1 to M do 3 extract shape-related features, f (I, s ), from the image using the current shape estimate; 4 apply the mth weak regressor to obtain the shape update δs using Eq. (3); 5 update the current shape estimate s ← s + δs; 6 end Algorithm 1: Cascaded shape regression for facial landmark localisation.…”

Section: Cascaded Shape Regressionmentioning

confidence: 99%

Face Detection, Bounding Box Aggregation and Pose Estimation for Robust Facial Landmark Localisation in the Wild

Feng

Kittler

Awais

et al. 2017

2017 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)

Self Cite

View full text Add to dashboard Cite

We present a framework for robust face detection and landmark localisation of faces in the wild, which has been evaluated as part of 'the 2nd Facial Landmark Localisation Competition'. The framework has four stages: face detection, bounding box aggregation, pose estimation and landmark localisation. To achieve a high detection rate, we use two publicly available CNN-based face detectors and two proprietary detectors. We aggregate the detected face bounding boxes of each input image to reduce false positives and improve face detection accuracy. A cascaded shape regressor, trained using faces with a variety of pose variations, is then employed for pose estimation and image pre-processing. Last, we train the final cascaded shape regressor for fine-grained landmark localisation, using a large number of training samples with limited pose variations. The experimental results obtained on the 300W and Menpo benchmarks demonstrate the superiority of our framework over state-of-the-art methods.

show abstract

“…Because the LqfNet is not specifically designed to handle misdetections, the descriptor distance d is weighted by the detection confidence c to shape the final recognition score s = c/d. UCCS a : The UCCS contribution relies on features from the publicly available 6 VGG Face descriptor network [24], which are extracted of 224×224 pixel cropped images. The enrollment is based on the Extreme Value Machine (EVM) [28], which is particularly designed for open-set recognition.…”

Section: Face Recognitionmentioning

confidence: 99%

“…Usually, a similarity between model and probe image is thresholded, where the threshold is computed based on a desired false acceptance rate. Other challenges included more difficult data, such as the Point and Shoot Challenge [6] or the Face Recognition Evaluation in Mobile Environment [11].…”

Section: Introductionmentioning

confidence: 99%

Unconstrained Face Detection and Open-Set Face Recognition Challenge

Günther

Herrmann

et al. 2017

2017 IEEE International Joint Conference on Biometrics (IJCB)

View full text Add to dashboard Cite

Face detection and recognition benchmarks have shifted toward more difficult environments. The challenge presented in this paper addresses the next step in the direction of automatic detection and identification of people from outdoor surveillance cameras. While face detection has shown remarkable success in images collected from the web, surveillance cameras include more diverse occlusions, poses, weather conditions and image blur. Although face verification or closed-set face identification have surpassed human capabilities on some datasets, open-set identification is much more complex as it needs to reject both unknown identities and false accepts from the face detector. We show that unconstrained face detection can approach high detection rates albeit with moderate false accept rates.By contrast, open-set face recognition is currently weak and requires much more attention.

show abstract

Report on the FG 2015 Video Person Recognition Evaluation

Abstract: Abstract-This report presents results from the Video Person Recognition Evaluation held in conjunction with the 11th

Cited by 41 publications

References 27 publications

A Comprehensive Survey on Pose-Invariant Face Recognition

A Comprehensive Survey on Pose-Invariant Face Recognition

Face Detection, Bounding Box Aggregation and Pose Estimation for Robust Facial Landmark Localisation in the Wild

Unconstrained Face Detection and Open-Set Face Recognition Challenge

Contact Info

Product

Resources

About