Skip to content

Latest commit

 

History

History
12 lines (7 loc) · 1.21 KB

README.md

File metadata and controls

12 lines (7 loc) · 1.21 KB

emo-detective

Preamble: This notebook was adapted from an Introduction to Machine Learning course assignment. This notebook is intended to demonstrate a basic understanding of convolutional neural network architecture.


Goal: To develop an emotion detection CNN-based model that can detect emotions from images in real-time.

Instructions: The entire full notebook should run less than 35 minutes to run on T4 GPU. The solutions for this notebook are self-contained. This notebook was intended to be executed and tested on Google Colab. Using the commands: runtimerun all.

Rationale: There are some general understanding questions and interpretations of the model placed in markdown cells with this emoji (❓) and answers placed in cells with this emoji (✅). This will provide some context on the application and an explanation of some of the design decisions made.

Dataset: Facial Expression Recognition 2013 Dataset. The required dataset is located within the FER-2013.rar file. Download it and store it in Google Drive. It should be under My Drive. Note: After uploading it, check the location by right-clicking on the file File InformationDetailsLocation.