Face Detection Using SURF Cascade

Jianguo Li, Tao Wang, Yimin Zhang Intel Labs China

Outline • Cascade Detection Revisited – Problems & motivations

• SURF-Cascade – SURF Feature – Maximizing AUC

• Benchmark • Conclusion

Cascade Detector Revisited • Five ingredients – Feature representation: Haar, HoG, … • Integral image to speedup feature extraction – Weak classifiers: dtree, linear SVM, … – Training algorithm: Boosting, … – Cascade structure: hard, soft, chain, … – Scan strategy: slide-window, …

∑f i

1i

( x ) > θ1

∑f i

2i

(x ) > θ 2

∑f i

ni

(x) > θ n

Problems & motivations • Practical detector requires ~1e-6 FPPW – Huge training set required – need scan >10^8 negative samples

• Large feature pool – In Haar cascade, > 200,000 features for 20x20 template.

• Slow convergence speed – Training based on two conflicted objectives: TPR/FPR • i.e, in each stage, set minTPR (0.995) and maxFPR (0.5) – Reach FPR=0.5 is easy in early stages – But TPR is not converged simultaneously – 1e-6=0.5^20, while 0.995^20=0.905

Weeks => Days => Hours?

SURF Cascade (1) • Features: SURF – 2x2 cell of patch – Each cell is 8-dim vector • Sum of dx, |dx| when dy >=0 • Sum of dx, |dx| when dy <0 • Sum of dy, |dy| when dx >=0 • Sum of dy, |dy| when dx <0 – Total is 2x2x8 = 32 dim feature vector – 8-channel integral images

• Feature Pool – – – – –

In a 40x40 face detection template Slide the patch (x, y, w, h) with fixed step = 4 pixels Each cell at least 8x8 pixels, w or h at least 16 pixels with 1:1, 1:2, 2:3… aspect-ratio (w/h) Totally 396 local SURF patches

• Weak classifier: logistic regression on 32dim SURF – h(x) = P(y|x, w) = 1/(1+exp(-ywx).

SURF Cascade (2) • Cascade training – AdaBoost in each stage

– Feature selection: maximize AUC score J

– Convergence test: AUC – Determine threshold when converged • Search on ROC curve with given TPR

Searching on ROC curve • In comparison, the Viola-Jones framework – Overall FPR 1e-6 = 0.5^20 – One stage TPR=0.995, overall 0.995^20 = 0.905

• Given TPR while FPR is adaptive – The FPR on 8-stage may like: • 1e-6 = 0.305x0.226x0.147x0.117x0.045x0.095x0.219x0.268

– Overall TPR = 0.995^8 = 0.970

Training performance • Implement in C/C++ on X86 – Parallelize the feature search step using OpenMP – SIMD for classifier (wx) and feature extraction

• Training dataset – 13000 faces from GENKI/FaceTracer database • With mirrors and resampling to obtain 39000 faces in total – 18000 non-face images from caltech101, image-net, etc.

• Training status – Platform: Intel Core-i7, 3.2GHz, 4-core, 8-thread. – On demand search of negative from non-face images • Totally scanned 13.6 billions of negative samples – Reach 1e-6 FPPW in 8-stages

Cascade statistics

#stages

#weak

Model-size

Hit-rate

training-time

(CMU Frontal)

on Core i7

VJ (OpenCV)

24

2912

>1MB

76.1%

~3 days

SURF

8

334

58KB

90.8%

47min

What if? • OpenCV Haar-training on the same dataset – Need 3 days (OpenMP tuned on)

• VJ’s criteria (TPR + FPR) for SURF? – Need 5 hours to reach 1e-6 at the 19-th stage

Evaluation on CMU+MIT frontal-set

Evaluation on UMass FDDB (frontal)

Multi-view SURF cascade on UMass

Some detection results

Detection speed • Test on three videos – A,B,C --- B has more faces than C in average

Intel Atom 1.6GHz

Intel Core i7 3.2GHz

• Why SURF cascade is faster than Haar-cascade – Average number of weak classifiers evaluated • SURF-cascade: 1.5 • Haar: 28 – Easy SIMD for SURF-cascade • 32-dim float => 128bit SIMD, 4-data in parallel • 1.5*32/4 = 12

Conclusion • Contributions – Introduce SURF feature for fast face detection – Propose AUC as single criterion for cascade training – Build a cascade face detector from billions of samples on PC within one hour.

• Advantages of SURF cascade – Very short cascade and small size (8 stages, ~58KB) – Accuracy is comparable to stage-of-the-art detectors. – Even faster than OpenCV optimized Haar-cascade

Thanks!

Face Detection Using SURF Cascade

Face Detection Using SURF Cascade. Jianguo Li, Tao Wang, Yimin Zhang ... 13000 faces from GENKI/FaceTracer database. • With mirrors and resampling to ...

423KB Sizes 0 Downloads 327 Views

Recommend Documents

Face Detection using SURF Cascade
rate) for the detection-error tradeoff. Although some re- searches introduced intermediate tuning of cascade thresh- old with some optimization methods [35, 2, ...

Fast Pedestrian Detection Using a Cascade of Boosted ...
on pedestrian detection using state-of-the-art locally extracted fea- tures (e.g. ... meaningful features if there is a large variation in object's ap- pearance .... The final strong classifier can be .... simple nonpedestrian patterns in the early s

SURF-Face: Face Recognition Under Viewpoint ...
A grid-based and dense extraction of local features in combination with a block-based matching ... Usually, a main drawback of an interest point based fea- .... navente at the Computer Vision Center of the University of Barcelona [13].

SURF-Face: Face Recognition Under Viewpoint ...
Human Language Technology and ..... In CVPR, Miami, FL, USA, June. 2009. ... In International Conference on Automatic Face and Gesture Recognition,. 2002.

Learning SURF Cascade for Fast and Accurate Object ...
ever, big data make the learning a critical bottleneck. This is especially true for training object detectors [37, 23, 41]. As is known, the training is usually required ...

Face Detection and Tracking Using Live Video Acquisition - MATLAB ...
Face Detection and Tracking Using Live Video Acquisition - MATLAB & Simulink Example.pdf. Face Detection and Tracking Using Live Video Acquisition ...

Face Detection Using Skin Likelihood for Digital Video ...
This project is based on self-organizing mixture network (SOMN) & skin color ... develop accurate and robust models for image data, then use the Gaussian ...

Face Detection Methods: A Survey
IJRIT International Journal of Research in Information Technology, Volume 1, Issue 11, November, 2013, Pg. 282-289 ... 1Student, Vishwakarma Institute of Technology, Pune University. Pune .... At the highest level, all possible face candidates are fo

Extraction Of Head And Face Boundaries For Face Detection ieee.pdf
Loading… Whoops! There was a problem loading more pages. Whoops! There was a problem previewing this document. Retrying... Download. Connect more apps... Extraction Of ... ction ieee.pdf. Extraction Of H ... ection ieee.pdf. Open. Extract. Open wit

Face Detection Algorithm based on Skin Detection ...
systems that analyze the information contained in face images ... characteristics such as skin color, whose analysis turns out to ..... software version 7.11.0.584.

Rapid Face Recognition Using Hashing
cal analysis on the recognition rate of the proposed hashing approach. Experiments ... of the images as well as the large number of training data. Typically, face ...

Using a Cascade of Asymmetric Resonators ... - Research at Google
with more conventional sound-analysis approaches. We use a ... ear extensions of conventional digital filter stages, and runs fast due ... nonuniform distributed system. The stage ..... the design and parameter fitting of auditory filter models, and 

Rapid Face Recognition Using Hashing
cal analysis on the recognition rate of the proposed hashing approach. Experiments ... of the images as well as the large number of training data. Typically, face ...

Face Recognition Using Eigenface Approach
which the person is classified by comparing its position in eigenface space with the position of known individuals [1]. The advantage of this approach over other face recognition systems is in its simplicity, speed and .... The experiment was conduct

Temporal Generalizability of Face-Based Affect Detection in Noisy ...
Department of Educational Psychology and Learning Systems4, Florida State .... with technology [1] and from observing students during pilot data collection (these ..... 171–182. Springer, Berlin. Heidelberg (1994). 25. Holmes, G., Donkin, A., ...

pdf-0738\face-detection-and-recognition-on-mobile-devices-by ...
pdf-0738\face-detection-and-recognition-on-mobile-devices-by-haowei-liu.pdf. pdf-0738\face-detection-and-recognition-on-mobile-devices-by-haowei-liu.pdf.

Quantitative Measurement of Face Detection Algorithm ...
Aug 5, 2008 - Quantitative Measurement of FD Algorithm Performance .... Speed. FERET. 735. 90.63%. 9.27%. 0.0%. 0.28 detik. DWI. 347. 89.78%. 10.22%.

Temporal Generalizability of Face-Based Affect Detection in Noisy ...
Cameras are a ubiquitous part of modern computers, from tablets to laptops to .... tures, and rapid movements can all cause face registration errors; these issues ...

Face detection advantage: Capture or odd-one out
attention. The participants' task was to search for the name of a politician or pop star among a list of letter strings while ignoring a flanking face distractor. They manipulated the perceptual load of the search task by varying the number of string

LNCS 4233 - Fast Learning for Statistical Face Detection - Springer Link
Department of Computer Science and Engineering, Shanghai Jiao Tong University,. 1954 Hua Shan Road, Shanghai ... SNoW (sparse network of winnows) face detection system by Yang et al. [20] is a sparse network of linear ..... International Journal of C

Support vector machine based multi-view face detection and recognition
theless, a new problem is normally introduced in these view- ...... Face Recognition, World Scientific Publishing and Imperial College. Press, 2000. [9] S. Gong ...

Credit Card Fraud Detection Using Neural Network
some of the techniques used for creating false and counterfeit cards. ..... The illustration merges ... Neural network is a latest technique that is being used in.