Partial Face Detection using Regions with Convolutional Neural Networks

Singh, Anjali (2015) Partial Face Detection using Regions with Convolutional Neural Networks. Masters thesis, Indian Institute of Technology Hyderabad.

CS13M1013.pdf - Submitted Version

Download (6MB) | Preview


Although many methods have been developed for holistic face detection, detecting partial faces hasn't been a successful endeavor yet. Partial faces frequently appear in real world environments like in surveillance videos and are quite difficult to detect. Recently, CNNs have shown very promising results with object detection in PASCAL VOC challenges. We propose an approach to detect partial faces along with holistic faces (frontal and profile views) present in natural scenarios. In the proposed method, we use CNN for feature extraction and representation. The drawback of CNN being computationally expensive is dealt with Selective Search using Segmentation, which reduces the search space to a great extent. We used FDDB Benchmarking for evaluating our method and the results were near the best results of all recent methods in the face detection domain.

[error in script]
IITH Creators:
IITH CreatorsORCiD
Item Type: Thesis (Masters)
Uncontrolled Keywords: convolutional Neural Networks, Deep learning, Partial Face Detection, Selective Search, TD413
Subjects: Computer science > Big Data Analytics
Divisions: Department of Computer Science & Engineering
Depositing User: Library Staff
Date Deposited: 31 Jul 2015 05:34
Last Modified: 14 May 2019 11:02
Publisher URL:
Related URLs:

Actions (login required)

View Item View Item
Statistics for RAIITH ePrint 1732 Statistics for this ePrint Item