Name
Abstract | ||
|---|---|---|
Person Identification of individuals has dependably been a challenge particularly when it needs to manage the big data sets and the robustness against the components influencing authentication, for example, posture variety, subject to camera distance, light variation, low-quality images and so on. Thus deep learning ends up being an awesome solution to conquer the above problems. Along these lines, we have outlined a design to recognize individuals by intertwining their gait and face biometric qualities utilizing a Deep Convolution Neural Network. In our work, the idea of Gait Energy Images (GEIs) is utilized to characterize human gait. From that point onward, both the vectors are combined, and the yield is given to the DCNN model for extracting features and classifying images. The proposed DCNN architecture is tried upon the publicly available CASIA Gait Dataset B, ORL Face Dataset, and FEI Face Database and a maximum identification percentage of 98.75% is accomplished on one test dataset and 97.50% accuracy on another dataset. We got improved results on Salt and Pepper noise, Gaussian Noise and Speckle Noise than previous work done in this field. |
Year | DOI | Venue |
|---|---|---|
2022 | 10.1007/s11277-022-09556-7 | Wireless Personal Communications |
Keywords | DocType | Volume |
Deep convolution neural network (DCNN), Gait energy images (GEIs), Gait recognition, Face recognition, Salt and pepper, Gaussian noise | Journal | 125 |
Issue | ISSN | Citations |
1 | 0929-6212 | 0 |
PageRank | References | Authors |
0.34 | 6 | 6 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Ankit Sharma | 1 | 0 | 0.34 |
| Neeru Jindal | 2 | 0 | 0.68 |
| Abhishek Thakur | 3 | 0 | 0.34 |
| Prashant Singh Rana | 4 | 0 | 0.68 |
| Bharat Garg | 5 | 0 | 0.68 |
| Rajesh Mehta | 6 | 0 | 0.68 |