Sorry, you need to enable JavaScript to visit this website.

IMPROVING CROSS-DATASET PERFORMANCE OF FACE PRESENTATION ATTACK DETECTION SYSTEMS USING FACE RECOGNITION DATASETS

Citation Author(s):
Sushil Bhattacharjee, Sebastien Marcel
Submitted by:
Amir Mohammadi
Last updated:
15 May 2020 - 10:22am
Document Type:
Presentation Slides
Document Year:
2020
Event:
Presenters Name:
Amir Mohammadi
Paper Code:
5846
Categories:

Abstract 

Abstract: 

Presentation attack detection (PAD) is now considered critically important for any face-recognition (FR) based access-control system. Current deep-learning based PAD systems show excellent performance when they are tested in intra-dataset scenarios. Under cross-dataset evaluation the performance of these PAD systems drops significantly. This lack of generalization is attributed to domain-shift. Here, we propose a novel PAD method that leverages the large variability present in FR datasets to induce invariance to factors that cause domain-shift. Evaluation of the proposed method on several datasets, including datasets collected using mobile devices, shows performance improvements in cross-dataset evaluations.

up
0 users have voted:

Dataset Files

autoencoder_error_icassp_2020_slides.pdf

(141)