人臉偵測(Face Detection)主要是在任意圖像中找到人臉的位置和大小的技術。早期人臉偵測演算法受到了燈光及遮蔽物影響其偵測準確率，隨著深度神經網路的發展已可以解決人臉偵測遇到的多數問題，並且獲得穩定的準確率，但是深度神經網路往往需要強大的硬體設備才能完成，昂貴的硬體設備並不符合許多實務的需求。單板電腦雖然價格低廉能夠大幅降低人臉偵測的門檻，但是單板電腦僅提供入門級的硬體設備，硬體的計算能力遠遠不及一般的電腦。
本論文的研究目的是提出應用於低成本的單板電腦之有效人臉偵測網路設計，以降低人臉偵測硬體設備需求，減少計算量，並且維持人臉偵測的準確率。設計的內容主要包含使用模型權重量化，預選框設計、訓練集篩選，最後透過樹莓派3B來實現人臉偵測。實驗結果顯示本論文所提出的方法，除了能夠有效降低偵測時間外，並能增加2.7%的平均準確率，實際偵測一張影像的時間只需要0.3秒，達到近似及時人臉偵測的目的，並且成功地利用低成本的設備準確且快速的偵測人臉的所在位置。

Face detection is a computer technology that identifies human faces in digital images. Light and facial occlusion affect the accuracy of face detection. Most face detection problems can be solved by deep learning model to get high accuracy. However, the use of deep learning model generally requires expensive and powerful devices. Single board computer is a cheaper device for face detection. Unfortunately, Single board computer is a less powerful device than general computers.
In this study, we proposed a face detection model design for low-cost single board computers so that it can reduce computing power and increase detection accuracy. The detection model design uses the quantized model, a set of modified anchor boxes, the dataset analysis. The overall design is implemented on the Raspberry Pi 3B. The experimental results verify that our method can increases the average accuracy by 2.7 %, and its cost time requires only 0.3sec. The proposed method can effectively detect faces with the use of the low-cost device.

[1] 人臉偵測參閱自維基百科全書。
https://en.wikipedia.org/wiki/Face_detection
[2] 深度學習解決方案透過GPU加速。NVIDIA。
https://www.nvidia.com/zh-tw/deep-learning-ai/solutions/.
[3] 2018年單板電腦Top 10。
https://www.eettaiwan.com/news/article/20180914NT01-10-Best-Single-Board-Computers-2018
[4] 針對不同應用需求 選擇不同構型單板電腦發展智能嵌入式應用。https://www.digitimes.com.tw/iot/article.asp?cat=130&id=0000397089_elo1dkma7s1enslvjjkp0
[5] 樹莓派(Raspberry Pi)，一款基於Linux的單晶片電腦。https://www.raspberrypi.org/
[6] M. Maksimović, V. Vujović, N. Davidović, V. Milošević, B. Perišić, “Raspberry Pi as Internet of Things hardware: Performances and Constraints,” International Conference on Electrical, Electronic and Computing Engineering, 2014
[7] N. Chowdhury, S. Nooman, S. Sarker, “Access Control of Door and Home Security by Raspberry Pi Through Internet,” International Journal of Scientific & Engineering Research, Nov. 2013.
[8] M. H. Yang, D.J. Kriegman,N. Ahuja, “Detecting faces in images: a survey,” IEEE Transactions on Pattern Analysis and Machine Intelligence, Nov. 2002.
[9] F. Schroff, D. Kalenichenko, J. Philbin “FaceNet: A Unified Embedding for Face Recognition and Clustering,” arXiv:1712.05877v1, Dec. 2017
[10] P. Viola, M. Jones, “Rapid Object Detection using a Boosted Cascade of Simple Features,” CVPR 2001
[11] K. Zhang, Z. Zhang, Z. Li, Y. Qiao, “Joint Face Detection and Alignment using Multi-task Cascaded Convolutional Networks,” arXiv:160402878, Apr. 2016
[12] W. Liu, D. Anguelov, D. Erhan, C. Szegedy, S. Reed, C.Y. Fu, A.C. Berg, “SSD: Single Shot MultiBox Detector,” arXiv:1512.02325v5, Dec. 2016
[13] Z.Q. Zhao, P. Z ,S.T Xu, X. Wu , “Object Detection with Deep Learning: A Review,” arXiv: 1807.05511v2, Apr. 2019
[14] S. Ren, K. He, R. Girshick, J. Sun , “Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks,” arXiv: 1506.01497v3, Jan. 2016
[15] chuanqi305 - Caffe implementation of Google MobileNet SSD detection network https://github.com/chuanqi305/MobileNet-SSD
[16] A. G. Howard, M. Zhu, B. Chen, D. Kalenichenko, W. Wang, T. Weyand, M. Andreetto, H. Adam, “MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications,” arXiv: 1704.04861v1, Apr. 2017
[17] K. Simonyan, A. Zisserman, “Very Deep Convolutional Networks for Large-Scale Image Recognition,” arXiv: 1409.1556v6, Apr. 2015
[18] S. Kligys, B. Chen, M. Zhu, M. Tang, A. Howard, H. Adam, D. Kalenichenko, “Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inference,” arXiv:1712.05877v1, Dec. 2017
[19] R. Krishnamoorthi, “Quantizing deep convolutional networks for efficient inference: A whitepaper,” arXiv:1806.08342v1, Jun. 2018
[20] 微軟LifeCam HD-3000 網路攝影機系列。
https://www.microsoft.com/accessories/zh-tw/products/webcams/lifecam-hd-3000/t3h-00014
[21] M. Abadi, A. Agarwal et al., “Tensorflow: Large-scale machine learning on heterogeneous distributed systems,” arXiv:1603.04467, Mar. 2016.
[22] TensorFlow Lite is a lightweight solution for mobile and embedded devices.
https://www.tensorflow.org/lite
[23] Tensorflow Object Detection API
https://github.com/tensorflow/models/tree/master/research/object_detection
[24] S. Yang, P. Luo, C. C. Loy, X. Tang, “WIDER FACE: A Face Detection Benchmark,” arXiv: 1511.06523v1, Nov. 2015
[25] T. Y. Lin, M. Maire, S. Belongie, L. Bourdev, R. Girshick, J. Hays, P. Perona, D. Ramanan, C. L. Zitnick, P. Dollár, “Microsoft COCO: Common Objects in Context,” arXiv: 1405.0312, Feb. 2015
[26] V. Jain, E. L. Miller, “FDDB: A Benchmark for Face Detection in Unconstrained Settings,” CVPR 2014