Open
Description
Task 2 : Image Preprocessing and Augmentation
This task is open to all. Your objective is to apply image preprocessing techniques and image augmentation to the provided dataset. Also you must make an 80-20 split to the train_set and create a validation_set. It is a folder which is to be present alongside test_set and train_set. You can use any preferred technique to do this.
Print the number of images in train_set and validation_set after making the split.
The notebook should include clean, well-commented code explaining what and why you are doing each step.
Important:
- Unclean code, rectified code in restricted regions, and uncommented code will not be merged. Comments must be detailed and explaining what and why of each step.
- Write clean, well-structured code to ensure better evaluation.
- There is a basic criteria of a number of processes and techniques that must be present in the code. It must be noted that doing just one thing of both pre-processing and augmentation will not make your PR merged.
Submission Guidelines
-
Colab Notebook Access:
- Open the provided
task2.ipynb
file in Google Colab. - Ensure that the Colab file link is restricted to view-only access for the following email IDs:
- Open the provided
-
Steps to Restrict Access:
- Click on the "Share" button in the top-right corner of the Colab notebook.
- Under "Get Link", select "Restricted" access.
- Add the specified email IDs with view-only permissions.
-
Submit the Colab Link:
- Create a plain text file (
rollno.txt
) and include the link to your Colab notebook. - Ensure the link is properly formatted.
- Create a plain text file (
-
Push the Link to the Repository:
- Push the text file (
rollno.txt
) to theTask2_solutions
folder in the repository as your submission for this task.
- Push the text file (
Resources Provided
- Notebook: Download the
task2.ipynb
file from the repository and open it in Google Colab for implementation.
Important Notes:
- Make sure the Colab link is restricted with view-only access for the specified email IDs only.
- Follow the provided notebook (
task2.ipynb
) for guidance on how to structure the task. - Ensure that your code is well-commented to explain the logic behind each preprocessing and augmentation step.
- Clean and structured code is essential for proper evaluation.
- For any queries join the discord channel and look for the channel named "Pestering-Data". Write your query in the discussion and wait for it to be resolved :)
- Is the Data really Pestering?