Schneiderman, H. and Kanade, T. Object Detection Using the Statistics of Parts, Viola, P. and Jones...

Post on 22-Dec-2015

218 views 0 download

Tags:

Transcript of Schneiderman, H. and Kanade, T. Object Detection Using the Statistics of Parts, Viola, P. and Jones...

Schneiderman, H. and Kanade, T. Object Detection Using the Statistics of Parts,

Viola, P. and Jones Robust Real-time Object Detection,

Andrew Gallagher

Dataset

• 16 images. 8 from my databases, and 8 from Flickr.

• Large ethnic, gender and age variety.

• 98 total faces. 66 “frontal” 32 profile.

Test

• The Viola Jones algorithm OpenCV implementation was used. (<2 sec per image).

• For Schneiderman Kanade, the www.pittpatt.com demo was used. (~10-15 seconds per image, including web transmission).

Results

• The Schneiderman-Kanade is very good and definitely out-performs Viola-Jones. (higher detection rate and lower FP simultaneously).

Ground Truth

OpenCV

FrontalDefault

OpenCV

ProfileFace

OpenCV Frontal Alt Tree

PittPatt

T=1

PittPatt

T=3

Frontal 66 40 20 36 59 55

Profile 32 9 3 3 23 10

False Positive

-- 24 7 2 22 0

Example Images

SchneidermanKanadeViola

Jones

Example Images

ViolaJones

SchneidermanKanade

Example Images

SchneidermanKanadeViola

Jones

SchneidermanKanadeViola

Jones

Example Images

ViolaJones

SchneidermanKanade