N
Nianfeng Liu
Researcher at Chinese Academy of Sciences
Publications - 8
Citations - 527
Nianfeng Liu is an academic researcher from Chinese Academy of Sciences. The author has contributed to research in topics: Iris recognition & Biometrics. The author has an hindex of 7, co-authored 7 publications receiving 407 citations.
Papers
More filters
Proceedings ArticleDOI
Accurate iris segmentation in non-cooperative environments using fully convolutional networks
TL;DR: Experimental results show that MFCNs are more robust than HCNNs to noises, and can greatly improve the current state-of-the-arts by 25.62% and 13.24% on the UBIRIS.v2 and CASIA.v4-distance databases, respectively.
Journal ArticleDOI
DeepIris: Learning Pairwise Filter Bank for Heterogeneous Iris Verification
TL;DR: DeepIris is a novel solution to iris recognition which learns relational features to measure the similarity between pairs of iris images based on convolutional neural networks, and EER of heterogeneous iris verification is reduced by 90% using DeepIris compared to traditional methods.
Proceedings ArticleDOI
LivDet iris 2017 — Iris liveness detection competition 2017
David Yambay,Benedict Becker,Naman Kohli,Daksha Yadav,Adam Czajka,Kevin W. Bowyer,Stephanie Schuckers,Richa Singh,Mayank Vatsa,Afzel Noore,Diego Gragnaniello,Carlo Sansone,Luisa Verdoliva,Lingxiao He,Yiwei Ru,Haiqing Li,Nianfeng Liu,Zhenan Sun,Tieniu Tan +18 more
TL;DR: Results of the third LivDet-Iris 2017 show that even with advances, printed iris attacks as well as patterned contacts lenses are still difficult for software-based systems to detect.
Proceedings ArticleDOI
Multi-patch convolution neural network for iris liveness detection
TL;DR: A Multi-patch Convolution Neural Network that automatically learns the features to detect hybrid pattern of fake iris images rather than handcraft to improve the robustness and accuracy for iris liveness detection.
Journal ArticleDOI
A Code-Level Approach to Heterogeneous Iris Recognition
TL;DR: Extensive experimental results of matching cross-sensor, high-resolution versus low-resolution and, clear versus blurred iris images demonstrate the code-level approach can achieve the highest accuracy in compared with the existing pixel-level, feature- level, and score-level solutions.