A cloud-based face video retrieval system with deep learning
- PDF / 2,713,955 Bytes
- 21 Pages / 439.37 x 666.142 pts Page_size
- 47 Downloads / 367 Views
A cloud‑based face video retrieval system with deep learning Feng‑Cheng Lin1 · Huu‑Huy Ngo1 · Chyi‑Ren Dow1
© Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract Face video retrieval is an attractive research topic in computer vision. However, it remains challenges to overcome because of the significant variation in pose changes, illumination conditions, occlusions, and facial expressions. In video content analysis, face recognition has been playing a vital role. Besides, deep neural networks are being actively studied, and deep learning models have been widely used for object detection, especially for face recognition. Therefore, this study proposes a cloudbased face video retrieval system with deep learning. First, a dataset is collected and pre-processed. To produce a useful dataset for the CNN models, blurry images are removed, and face alignment is implemented on the remaining images. Then the final dataset is constructed and used to pre-train the CNN models (VGGFace, ArcFace, and FaceNet) for face recognition. We compare the results of these three models and choose the most efficient one to develop the system. To implement a query, users can type in the name of a person. If the system detects a new person, it performs enrolling that person. Finally, the result is a list of images and time associated with those images. In addition, a system prototype is implemented to verify the feasibility of the proposed system. Experimental results demonstrate that this system outperforms in terms of recognition accuracy and computational time. Keywords Video retrieval · Face recognition · Cloud-based system · Deep learning · FaceNet
* Feng‑Cheng Lin [email protected] Huu‑Huy Ngo [email protected] Chyi‑Ren Dow [email protected] 1
Department of Information Engineering and Computer Science, Feng Chia University, Taichung, Taiwan
13
Vol.:(0123456789)
F.-C. Lin et al.
1 Introduction Multimedia information systems have been massively and diversely used in research and practical applications. Therefore, it is also very common to see multimedia items, in particular, videos in our daily lives. For example, distance learning, video conferencing and multimedia broadcasting. However, it is becoming difficult to acquire relevant video data and manage by human effort. Face video retrieval is an application that can search videos of a particular person over a video database when one face video of that person is given. Face video retrieval has been an appealing research topic in recent years and utilized in many fields. For example, video surveillance systems that can locate and recognize suspects from surveillance videos [19]. With the rapid development of social networking and semantic websites, video searching and video sharing become very popular. Besides, in the entertainment industry, especially movie industry, there has been an increasing demand for techniques that allow audiences to seek specific actors or actresses in a film and directors to look for the cast that best suits their mo
Data Loading...
