Live Face Verification with Multiple Instantialized Local Homographic Parameterization

Authors: Chen Lin, Zhouyingcheng Liao, Peng Zhou, Jianguo Hu, Bingbing Ni

IJCAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experimental results on several face benchmarks well demonstrate the superior performance of our method.4 Experiments Live face verification must be robust across different types of attacks. We evaluate our proposed approach on the OULUNPU database [Boulkenafet et al., 2017] and the Replay Attack database [Chingovska et al., ].
Researcher Affiliation Collaboration Chen Lin1, Zhouyingcheng Liao1, Peng Zhou1, Jianguo Hu2 and Bingbing Ni1 1Shanghai Jiao Tong University 2Minivision linchen040329@sjtu.edu.cn, patrickliao2007@gmail.com, zhoupengcv@sjtu.edu.cn, hujianguo@minivision.cn, nibingbing@sjtu.edu.cn
Pseudocode Yes Algorithm 1 Learning algorithm for multi-patch examination module
Open Source Code No The paper states that codes are built on Pytorch, but does not provide an explicit statement or link to their own open-source code for the described methodology.
Open Datasets Yes We evaluate our proposed approach on the OULUNPU database [Boulkenafet et al., 2017] and the Replay Attack database [Chingovska et al., ]. The OULU-NPU database contains 990 real face videos and 3960 attack face videos. And it is divided into training, development and testing subsets. The Replay Attack database contains 200 real face videos and 1000 attack videos. It is divided into training, development and testing subsets.
Dataset Splits Yes The OULU-NPU database contains 990 real face videos and 3960 attack face videos. And it is divided into training, development and testing subsets. It claims that the models should be learned on the training subset, fine-tuned on the development subset, and tested on the testing subset. The Replay Attack database contains 200 real face videos and 1000 attack videos. It is divided into training, development and testing subsets.
Hardware Specification Yes Our set up consists of a intel i5-5350 CPU, a 8GB Memory and a Logitech HD C920 Pro webcam.
Software Dependencies No The paper states 'All training and testing codes are built on Pytorch [Paszke et al., 2017]', but it does not specify a version number for Pytorch or any other software dependencies.
Experiment Setup Yes In our experiments, we use Stochastic Gradient Descent (SGD) to optimize our model. 8 videos are randomly chosen and adjacent two frames are sampled from each video to form a batch. The Res Net in our model is initialized from a pre-trained model for Image Net [Deng et al., 2009] classification. We start training with a learning rate of 0.005 and a weight decay of 0.001, and decrease the learning rate by 1/10 when the loss goes steady.