Comparing robustness of pairwise and multiclass neural-network systems for face recognition
| dc.contributor.author | Uglov, J. | en_GB |
| dc.contributor.author | Jakaite, Livija | en_GB |
| dc.contributor.author | Schetinin, Vitaly | en_GB |
| dc.contributor.author | Maple, Carsten | en_GB |
| dc.date.accessioned | 2013-05-30T13:09:44Z | |
| dc.date.available | 2013-05-30T13:09:44Z | |
| dc.date.issued | 2008 | |
| dc.identifier.citation | Uglov, J., Jakaite, L., Schetinin, V. & Maple, C. (2008) 'Comparing robustness of pairwise and multiclass neural-network systems for face recognition', EURASIP Journal on Advances in Signal Processing, 2008 pp.64. | en_GB |
| dc.identifier.issn | 1687-6180 | |
| dc.identifier.doi | 10.1155/2008/468693 | |
| dc.identifier.uri | http://hdl.handle.net/10547/293038 | |
| dc.description.abstract | Noise, corruptions, and variations in face images can seriously hurt the performance of face-recognition systems. To make these systems robust to noise and corruptions in image data, multiclass neural networks capable of learning from noisy data have been suggested. However on large face datasets such systems cannot provide the robustness at a high level. In this paper, we explore a pairwise neural-network system as an alternative approach to improve the robustness of face recognition. In our experiments, the pairwise recognition system is shown to outperform the multiclass-recognition system in terms of the predictive accuracy on the test face images. | |
| dc.language.iso | en | en |
| dc.relation.url | http://asp.eurasipjournals.com/content/2008/1/468693 | en_GB |
| dc.rights | Archived with thanks to EURASIP Journal on Advances in Signal Processing | en_GB |
| dc.subject | G760 Machine Learning | en_GB |
| dc.subject | face recognition systems | en_GB |
| dc.subject | face recognition | en_GB |
| dc.subject | neural networks | en_GB |
| dc.subject | multiclass-recognition system | en_GB |
| dc.subject | corruptions | en_GB |
| dc.subject | noise | en_GB |
| dc.title | Comparing robustness of pairwise and multiclass neural-network systems for face recognition | en |
| dc.type | Article | en |
| dc.identifier.journal | EURASIP Journal on Advances in Signal Processing | en_GB |
| html.description.abstract | Noise, corruptions, and variations in face images can seriously hurt the performance of face-recognition systems. To make these systems robust to noise and corruptions in image data, multiclass neural networks capable of learning from noisy data have been suggested. However on large face datasets such systems cannot provide the robustness at a high level. In this paper, we explore a pairwise neural-network system as an alternative approach to improve the robustness of face recognition. In our experiments, the pairwise recognition system is shown to outperform the multiclass-recognition system in terms of the predictive accuracy on the test face images. |
