This project aims to further the line of researches by predicting the genre and valence mood of the audio simultaneously by suing a multi-output CNNs to learn the features of mel spectrograms generated from the audio.

Notably, attention mechanism is applied in combination with CNN to extract the features of audio samples. The structure is shown below. For detailed model architexture, please refer to paper.

In this project, I build a classification model to predict pathological grade from medical imaging based on Inception-Resnet-v2.

Inception-Resnet-v2 is a CNN released by Google. It is part of the evolution of GoogLeNet, which is a remarkable achievemnet in the exploration of residual learning in inception networks.

The project is my bachelor thesis sponsored by Foxconn.

Facility Management Control System (FMCS) is a monitoring system for afactory′s cooling water, electricity, compressed air, etc. This system enables factory monitors to receive a real-time resource usage and gives them the essential information to adjust the resource input.

This project aims to implement an AI-enabled FMCS system that helps operators improve energy efficiency of factory facility equipments, such as cooling water, compressed dry air (CDA), fans and motor.

DJTable is all about having fun and enabling amazing experience for gamers to create their own music like a DJ.The goal of DJTable is to make a collaborative tabletop music experience for users through soothing music and synesthetic visual graphics.

For this project, I compare performances on multiple models such as CNN, Xgboost, logistic regression etc. with ensembling of text classification for Chinese. The main model I tested was CNN with great improvement combined with xgboost ensemble. However, ensemble of logistic regression did't help a lot.

The project is partitioned into several aspects. It mainly contained some key tasks related to computer vision in autonomous driving field.

It is expected to be finished by October.