We build an API for predicting the popularity of Mandopop/C-pop song. You can simply upload a song, and the model will tell you the estimated popularity of it. The model is based on 1. an inception CNN which takes mel-spectrogram of the song as input and 2. a fully-connected DNN which takes auto-tagging information as input.
Using ridge regression, we found memory-related neural activity in parietal cortex can successfully predict the memory content (which is vectorized by the final fully connected layer's output from VGG16.)
In this notebook, I am going to show you How to extract the features for a set of images at a certain layer of VGG16 pretrained model with PyTorch. How to use PCA(Principle component analysis) to reduce the feature dimension for better visualization.
Although the theory basis of my major, cognitive neuroscience, is based on psychology and biology, the skills I need to tackle with the research work of this field are most data science related.