An implementation of convolutional neural networks for writerless signature verification. Originally run using Google Colab. Created as a the output for the PSHS Research 1 subject.
cs_research_1.ipynb standardizes and compiles sigcomp datasets across different years, while cs_research_2_old.ipynb trains the convolutional neural network using tensorflow, to classify signatures as either genuine or fake.