Task Instruction: Hand Segmentation in Images (Background Removal)

Task Description

Your task is to develop a model or method for segmenting hands from images and produces accurate results. This means isolating the hand by removing the background. The dataset will be provided by Aitaca.

Requirements

1. Dataset: You should use this dataset for training and testing your model/method. Pay special attention to the accuracy when removing the background, avoiding noise in the contouring. CThe dataset can be accessed via the following Google Drive link: https://drive.google.com/file/d/1Mc72BLGnZZQ0mhnh9rHOdJzYmITN-PPL/view?usp=sharing. Make sure to download and use this dataset for your task. Model is supposed to remove the background as in the images given in the dataset.

2. Programming Language: You should use Python for this task. You can choose to work in a Python notebook (e.g., Jupyter Notebook) as well.

3. Methods: You are free to choose any methods you see fit to achieve accurate hand segmentation. This may include neural networks, image processing techniques, or a combination of both.

4. Submission:

   • Ensure that your code is organized and includes all necessary components for the task. If you use any data processing steps as part of your solution, include that code as well.
   • If you choose to work with neural networks, please submit the trained model along with the code used for training.
   • If you find and use pre-trained neural networks, include the details and sources of those networks.
   • Include a part in your code where you visualize the segmentation results on a subset of the dataset's images.

5. Code Documentation: Ensure that your code includes documentation, comments, and a separate README file that explains how to use your code and details the submitted code, including any specific dependencies or setup instructions. If your code relies on specific data folder structures or arrangements, please provide an explanation in the README file about how the data should be organized within the repository.

6. Submission Format: You must fork this respository and upload all your work to the fork repository.

7. Contact Information: If you have any questions or need clarification regarding the task, please contact tech@aitaca.io.

8. Confidentiality: Please ensure that you respect any confidentiality agreements or data usage restrictions related to the dataset provided.

We look forward to reviewing your work and assessing your skills as a Machine Learning Engineer. Good luck with the task!