Deep Neural networks have already surpassed human-level performance in tasks such as object recognition and detection. However, deep networks were lagging far behind in tasks like generating artistic artifacts having high perceptual quality until recent times. Creating better quality art using machine learning techniques is imperative for reaching human-like capabilities, as well as opens up a new spectrum of possibilities. And with the advancement of computer hardware as well as the proliferation of deep learning, deep learning is right now being used to create art. For example, an AI-generated art won’t be sold at an auction for a whopping $432,500.
Create a Python package for plug in and play😎 different image processing and deep learning tasks without having to know about its working and the scary math that follows 😱.
In the process of doing so, we want the participants to:
- Learn various concepts in Deep Learning and Computer Vision🐱🏍
- Implement them in the form of scripts 👩💻
- Reimplement using more optimized methods and do a comparative analysis 🕵️♀️
- Create your own datasets, test your scripts on them and make it available for the community to work further 🔨
- Build small APIs for each task (wherever applicable)
- Build web apps for certain tasks using TensorFlow.JS or Flask
- Build Flutter apps with TensorFlow lite for certain tasks 📲
- Increment your personal portfolio with certain tasks you do here as standalone projects 👩💼
- Work for applying the various tasks undertaken into several real world scenarios
- Write blog posts about your contributions (Medium: @techquilla) to share with the world ✏
- Learn to read and implement research papers 🔬
- Don't push anything to the root directory. Always use specific subdirectories inside the deeppixel directory
For each of the tasks:
- Experiment building this using a Jupyter notebook locally or on Google Colab
- Build a script for the task inside the
respectivedirectory - Inside the each directory, create two folders
inputandoutputto be used for the input and output images - Use argparse library so that the input image and output path can be given as arguments in the terminal while running the script
- Create a
requirements.txtfile and specify the modules used - Try your script/notebook with multiple images and store the results in output folder
- Comment your script/notebook well
- Create/Update the ReadME.MD file:
- Name of the task
- A small description
- Approach used
- Input and Output Images
- Make sure you have the model weight or any related files like haarcascades to run the script in the same direcorty
- Give a Pull Request
- In your PR, please try to give a link to a Colab Notebook(if applicable) as a comment.
- In your PR, put a reference to the issue it is for
- once your model/script works, curate your own dataset and get the outputs..(Try 10 images) adn give a second PR about your own Dataset
- Try out other methods to implement the above
__Please do not use a code from someone else's repo or a blog like PyImageSearch directly. You can definitely refer to others' code. But make sure you have some contributions of your own into it. If you strongly use a code from someone else,please credit them properly in the README file of the respective directory.
1. Fork this repository.
2. Clone the forked repository.
git clone https://github.com/<your-github-username>/DeepPixel
3. Navigate to the project directory.
cd DeepPixel
4. Create a new branch.
git checkout -b <your_branch_name>
5. Make changes in source code.
6. Commit your changes.
git add .
git commit -m "<your_commit_message>"
7. Push your local branch to the remote repository.
git push -u origin <your_branch_name>
8. Create a Pull Request!
Congratulations! You have just made your contribution to DeepPixel project.
- Python(Mandatory)
- Git(Mandatorty)
- Linux Command Line(Mandatory)
- Elementary Knowledge of Deep Learning or Computer Vision (Mandatory)
- Ability to use TensorFlow 2.0/PyTorch/Keras/fast.ai (any one is suggested)
- OpenCV(Required)
- HTML,CSS,JavaScript(can be picked up on the go)
- TensorFlow.JS
And above all the willingness to learn and contribute!
- PyImageSearch-One of the best places on the Internet for anything on Computer Vision
- Fast.ai-Deep Learning course+Forum for any level that you will fall in love with!
- MIT:Intro to Deep Learning - shows you more real world applications as you dive into deep learning
- CS231n: Convolutional Neural Networks for Visual Recognition
- MIT Deep Learning - From self driving cars to Alpha fold..they help you dig into the latest in the field..Lex Fridman's podcasts are just 🔥🔥
- Deep Learning Specialization - Most popular MOOCs that is used by thousands all over the world to get started with Deep Learning
We would love to hear from you! We communicate on the following platforms: