Detect and Classify Species of Fish from Fishing Vessels with Modern Object Detectors and Deep Convolutional Networks

Speaker:       Felix Yu
               Founder of Toppick Analytics

Title:   "Detect and Classify Species of Fish from Fishing Vessels
          with Modern Object Detectors and Deep Convolutional Networks"

Date:    Thursday, 25 May 2017

Time:    2:00 - 3:00pm

Venue:   Room 3501 (via lifts 25/26), HKUST

Abstract:

The goal of the Kaggle competition is to develop models to automatically
detect and classify species of fish that the fishing boats catch. This is
a very challenging problem since there were very few training samples
available, and that difference between different category of fishes can be
highly subtle. I will talk about my approach of using a 2 staged process
that utilizes state-of-the-art Convolutional Neural Networks and Object
Detectors to achieve the 8th place.


*******************
Biography:

Felix is the Founder of Toppick Analytics, a financial technology company
specializing in building financial data analytics products. Prior to
founding Toppick, Felix earned his Masters Degree in Computational and
Mathematical Engineering from Stanford University and Bachelor Degree in
Financial Engineering from Columbia University.

On the side, Felix is passionate in machine learning and had competed in
various data science/predictive analytics competitions. He achieved
"master" status in Kaggle, which is awarded to a group of top performers
in data science competitions hosted on Kaggle platform. Felix has
extensive experience using Python for web programming and data analytics.