Portfolio

• This Portfolio contains 3 projects described below.
• This portfolio is dedicated to image recognition, time series and recommendation systems.
• The short-term objective is to provide you with code on image recognition (Birds_Detection), time series (Ashare) and recommendation systems (MovieRecommenderProject) themes that you can use as inspiration for your projects.
• The medium term objective is to offer open source code for the bird recognition project and applicable to any image database to be classified.

Birds_Detection (added in October 2020)

The puropose is to detect a maximum of birds in fields. Farmers loose a part of their harverst because birds eat seeds in their fields. That is why we want to detect them in order to make them run away. In this regard Custom Convolutional Neural Networks such as Lenet, VGG16 or YoloV2 was used.

The picture below is an overview of the project where detected birds are represented with green squares.

If you want give a try with Yolo and detect birds on pictures I made available, you can type :

• cd Birds_Detection/Demonstration
• python3 inference.py If you would like to have more details about this project go to the README in Birds_Detection, I will explain the strategies I used.

Ashare (start in November 2019)

The goal is to develop models of metered building energy usage in the following areas:

• chilled water, electric,
• hot water,
• steam meters.
• The data comes from over 1,000 buildings over a three-year timeframe. It refers to the kaggle competition. For more details and to download data follow the link below:
• https://www.kaggle.com/c/ashrae-energy-prediction

MovieRecommenderProject (start in Februar 2019)

I made a recommender engine proposing to a user the best movie(s) corresponding to its tastes. You have the possibility to quickly try it by opening a terminal in MovieRecommenderProject/MovieRecommenderEngine and typing : python3 Action_Engine.py You should get something like that in your terminal :

Please find If you want more details in README in MovieRecommenderProject/MovieRecommenderEngine

If you want more details or to work with me, please feel free to contact me by email: marc.pozzo@gmail.com