International Journal of Innovative Research in Computer and Communication Engineering
ISSN Approved Journal | Impact factor: 8.771 | ESTD: 2013 | Follows UGC CARE Journal Norms and Guidelines
| Monthly, Peer-Reviewed, Refereed, Scholarly, Multidisciplinary and Open Access Journal | High Impact Factor 8.771 (Calculated by Google Scholar and Semantic Scholar | AI-Powered Research Tool | Indexing in all Major Database & Metadata, Citation Generator | Digital Object Identifier (DOI) |
| TITLE | Implementation of Hybrid Deep Learning-Based Face Recognition Using SIFT-CNN Integration |
|---|---|
| ABSTRACT | Face recognition under real-world variations in pose, illumination, and partial occlusion remains a key challenge in computer vision. This paper presents the detailed implementation of a hybrid face recognition system integrating four classical feature extraction techniques — Scale Invariant Feature Transform (SIFT), Histogram of Oriented Gradients (HOG), Gabor filters, and Canny edge detection — with a custom Convolutional Neural Network (CNN) architecture. The system is implemented and evaluated on two standard benchmark databases: the ORL face database (400 images, 40 subjects) and the Sheffield face database (564 images, 20 subjects). Multiple combinations of activation functions (Softmax, Sigmoid) and optimization algorithms (Adam, Adamax, RMSprop, SGD) are systematically evaluated. Experimental results demonstrate that the SIFT+CNN combination achieves superior recognition accuracy, reaching 100% on Sheffield with Adam+Softmax and 92.5% on ORL. The system outperforms all compared existing methods. Complete implementation details covering preprocessing, feature extraction with mathematical formulations, CNN architecture, training strategy, performance evaluation with accuracy/loss curves, and comparative analysis are presented. |
| AUTHOR | PANCHALWAR DIVYANI A., WAKHARE AMOL D. M.Tech Student, Department of CSE, Deogiri Institute of Engineering and Management Studies, Chh. Sambhajinagar, Maharashtra, India Assistant Professor, Department of CSE, Deogiri Institute of Engineering and Management Studies, Chh. Sambhajinagar, Maharashtra, India |
| VOLUME | 184 |
| DOI | DOI: 10.15680/IJIRCCE.2026.1405102 |
| pdf/102_Implementation of Hybrid Deep Learning-Based Face Recognition Using SIFT-CNN Integration.pdf | |
| KEYWORDS | |
| References | [1] K. Zhang et al., "Joint face detection using MTCNN," IEEE Signal Process. Lett., vol. 23, pp. 1499-1503, 2016. [2] A. Learned-Miller et al., "Labeled faces in the wild: A survey," Adv. Face Detection, pp. 189-248, 2016. [3] W. Zhao et al., "Face recognition: A literature survey," ACM Comput. Surv., vol. 35, pp. 399-458, 2003. [4] Y. LeCun, Y. Bengio, and G. Hinton, "Deep learning," Nature, vol. 521, pp. 436-444, 2015. [5] J. Deng et al., "ArcFace: Additive angular margin loss," Proc. IEEE CVPR, pp. 4690-4699, 2019. [6] D. G. Lowe, "Distinctive image features from scale-invariant keypoints," IJCV, vol. 60, pp. 91-110, 2004. [7] A. Kumar et al., "Face detection techniques: A review," Artif. Intell. Rev., vol. 52, pp. 927-948, 2019. [8] L. Chen and M. Williams, "Enhanced SIFT-based face recognition," Pattern Recognit. Lett., vol. 128, pp. 243-251, 2019. [9] N. Dalal and B. Triggs, "HOG for human detection," Proc. IEEE CVPR, pp. 886-893, 2005. [10] X. Zhang et al., "Automatic texture synthesis for face recognition," Pattern Recognition, vol. 89, pp. 61-71, 2019. [11] M. Rodriguez and J. Smith, "Multi-scale HOG descriptors," IEEE Trans. Inf. Forensics Secur., vol. 14, pp. 2090-2102, 2019. [12] C. Liu et al., "Gabor feature based classification," IEEE Trans. Image Process., vol. 11, pp. 467-476, 2020. [13] H. Wang and S. Chen, "Adaptive Gabor filter selection," Comput. Vis. Image Underst., vol. 195, pp. 102-114, 2020. [14] Y. Taigman et al., "DeepFace," Proc. IEEE CVPR, pp. 1701-1708, 2014. [15] F. Schroff et al., "FaceNet," Proc. IEEE CVPR, pp. 815-823, 2015. [16] Y. Sun et al., "DeepID3," arXiv:1502.00873, 2015. [17] Y. Wen et al., "Center loss for deep face recognition," Proc. ECCV, pp. 499-515, 2016. [18] K. Zhang and Y. Liu, "SIFT-CNN integration," IEEE Trans. NNLS, vol. 31, pp. 2456-2468, 2020. [19] R. Johnson et al., "Multi-scale SIFT-CNN fusion," Pattern Recognition, vol. 98, pp. 107-118, 2020. [20] C. Martinez and B. Thompson, "HOG-CNN fusion," CVIU, vol. 201, pp. 103-115, 2020. [21] S. Lee et al., "Adaptive HOG-CNN systems," IEEE Trans. Image Process., vol. 29, pp. 8456-8467, 2020. [22] L. Wang et al., "Gabor-CNN integration," Neural Networks, vol. 134, pp. 78-89, 2021. [23] H. Chen and M. Rodriguez, "Multi-orientation Gabor-CNN," Pattern Recognit. Lett., vol. 142, pp. 156-164, 2021. [32] J. Canny, "Computational approach to edge detection," IEEE TPAMI, vol. 8, pp. 679-698, 1986. |
