basel-ay / face-landmark-detection

Face-Landmark-Detection

Introduction

Face Detection Systems have great uses in today’s world which demands security, accessibility or joy! Today, we will be building a model that can plot 15 key points on a face.

Face Landmark Detection models form various features we see in social media apps. The face filters you find on Instagram are a common use case. The algorithm aligns the mask on the image keeping the face landmarks as base points.

In this notebook, we'll develop a model which marks keypoints on a given image of a human face. We'll build a Convolutional Neural Network which takes an image and returns a array of keypoints.

We'll require a GPU Hardware accelerator for training the model. Change the runtime type to GPU by going to Tools > Change Runtime Type > Hardware Accelerator > GPU.

Table of Contents

1.Introduction

2.Approach to the problem statement

3.Data loading and Preprocessing

4.Model building

5.Model training

6.Evaluation

Dataset : https://www.kaggle.com/drgilermo/face-images-with-marked-landmark-points

Notebook on Kaggle : https://www.kaggle.com/basel99/face-landmark-detection