Fine-grained face verification: FGLFW database, baselines, and human-DCMN partnership
作者：
Highlights：
• New fine-grained face verification task to detect the imposter who deliberately seek a people with similarly-looking face.
• Fine-Grained LFW (FGLFW) database that evaluates the new fine-grained face verification task.
• Controlled human survey reports 99.85% accuracy on LFW, but only 92.03% accuracy on FGLFW.
• Deep Convolutional Maxout Network (DCMN) that outperforms current techniques such as Deepface, DeepID2, and VGG-Face on LFW and FGLFW experiments.
• The finding that human and DCMN are highly complementary on the fine-grained face verification task.
摘要
•New fine-grained face verification task to detect the imposter who deliberately seek a people with similarly-looking face.•Fine-Grained LFW (FGLFW) database that evaluates the new fine-grained face verification task.•Controlled human survey reports 99.85% accuracy on LFW, but only 92.03% accuracy on FGLFW.•Deep Convolutional Maxout Network (DCMN) that outperforms current techniques such as Deepface, DeepID2, and VGG-Face on LFW and FGLFW experiments.•The finding that human and DCMN are highly complementary on the fine-grained face verification task.
论文关键词：Fine-grained visual recognition,Face verification,Labeled face in the wild,Deep learning
论文评审过程：Received 16 July 2016, Revised 27 November 2016, Accepted 28 November 2016, Available online 29 November 2016, Version of Record 12 March 2017.
论文官网地址：https://doi.org/10.1016/j.patcog.2016.11.023
