Eye Movement Detection using DL

[abhisheks008]

Deep Learning Simplified Repository (Proposing new issue)

:red_circle: Project Title : Eye Movement Detection using DL :red_circle: Aim : Create an DL project which will identify the movements of the eyeballs. :red_circle: Dataset : https://www.kaggle.com/datasets/priyankraval/eyet4empathy-eye-movement-and-empathy-dataset :red_circle: Approach : Try to use 3-4 algorithms to implement the models and compare all the algorithms to find out the best fitted algorithm for the model by checking the accuracy scores. Also do not forget to do a exploratory data analysis before creating any model.

---

📍 Follow the Guidelines to Contribute in the Project :

• You need to create a separate folder named as the Project Title.
• Inside that folder, there will be four main components.
  • Images - To store the required images.
  • Dataset - To store the dataset or, information/source about the dataset.
  • Model - To store the machine learning model you've created using the dataset.
  • requirements.txt - This file will contain the required packages/libraries to run the project in other machines.
• Inside the Model folder, the README.md file must be filled up properly, with proper visualizations and conclusions.

---

:red_circle::yellow_circle: Points to Note :

• The issues will be assigned on a first come first serve basis, 1 Issue == 1 PR.
• "Issue Title" and "PR Title should be the same. Include issue number along with it.
• Follow Contributing Guidelines & Code of Conduct before start Contributing.

---

:white_check_mark: To be Mentioned while taking the issue :

• Full name :
• GitHub Profile Link :
• Email ID :
• Participant ID (if applicable):
• Approach for this Project :
• What is your participant role? (Mention the Open Source program)

---

Happy Contributing 🚀

All the best. Enjoy your open source journey ahead. 😎

[Rani1303]

Hey @abhisheks008 , I am interested in this can you assign this to me under SWOC .

[abhisheks008]

Wait for the program to officially start. Issues will be assigned to you after the orientation. @Rani1303

[Rani1303]

Okay

[dsid271]

Full name : Sidhartha Dondapati GitHub Profile Link : https://github.com/dsid271 Email ID : [email protected] Participant ID (if applicable): Approach for this Project : VGG16 or LeNet-5 What is your participant role? (Mention the Open Source program): SWOC 2024

[abhisheks008]

Hi @Rani1303 and @dsid271 please share your approach for solving this issue. What are models you are planning to implement, do let me know here, based on that this issue will be assigned.

[dsid271]

I plan to use openCV assuming that the eye movement is being recorded in real-time

[abhisheks008]

I plan to use openCV assuming that the eye movement is being recorded in real-time

What are the other models you wanna work on? As you need to implement at least 2-3 models to make a comparison and find out the best fitted model for this dataset and project.

[dsid271]

Sorry for the confusion, I would like to work on CNN approach I will try to compare VGG and LeNet

[abhisheks008]

Sorry for the confusion, I would like to work on CNN approach I will try to compare VGG and LeNet

Cool. That's a good approach.

[Rani1303]

Hey! @abhisheks008 I will be using CNN, RNN and CapsNet for the model comparison. And if required will work on GoogLeNet also.

[abhisheks008]

Based on the approaches you guys have proposed, I will go with @Rani1303. Use CNN, RNN, CapsNet and also GoogLeNet for this project and find out the best fitted model for this project.

@dsid271 there are lots of open issues present here. You can check out any of them.

Issue assigned to @Rani1303. You can start working on it.

[Rani1303]

Thanks @abhisheks008 will work on this issue.

[veer-chheda]

Full Name: Veer Chheda GitHub profile: https://github.com/veer-chheda Email ID: [email protected] Participant ID: N/A Approach: I'd like to utilise OpenCV and albumentations for data augmentation to improve accuracy; moreover, I'd use a CNN architecture as the problem is computer vision-based. I'd use convolutional layer weights from VGG16, MobileNet, and ResNet and fine tune the last layer weights according to the dataset provided to improve the accuracy. Participant Role: Contributor (GSSoC'24) @abhisheks008

[abhisheks008]

Issue assigned to you @veer-chheda. You can start working on it.

[veer-chheda]

@abhisheks008 Am I free to use custom datasets with images used for detection using OpenCV or I have to use the one provided?

[abhisheks008]

It's better to use the given one. If you want to use other datasets, please share the dataset URL here, let me check that.

[veer-chheda]

The dataset provided is for a regression problem to calculate the empathy score, whereas the project focuses on eye movement or gaze detection, which can be done with OpenCV and MediaPipe and without the need for an as such dataset.

[abhisheks008]

The dataset provided is for a regression problem to calculate the empathy score, whereas the project focuses on eye movement or gaze detection, which can be done with OpenCV and MediaPipe and without the need for an as such dataset.

Hi @veer-chheda sorry for being late. Can you please share the dataset URL here? Are you fetching the dataset from Kaggle?

[veer-chheda]

I am going to use Google MediaPipe for this project - https://developers.google.com/mediapipe, for eye movement detection as this would require no need for a dataset and give a better accuracy. Let me know if I can go ahead with the project.

[abhisheks008]

Cool, then in the Dataset folder create a README file and mention the source of the dataset along with a short brief of the dataset. I hope this will be good to go for you too. @veer-chheda