Welcome to the Convolutional Neural Networks (CNN) project in the AI Nanodegree! In this project, you will learn how to build a pipeline that can be used within a web or mobile app to process real-world, user-supplied images. Given an image of a dog, your algorithm will identify an estimate of the canine’s breed. If supplied an image of a human, the code will identify the resembling dog breed.
Along with exploring state-of-the-art CNN models for classification and localization, you will make important design decisions about the user experience for your app. Our goal is that by completing this lab, you understand the challenges involved in piecing together a series of models designed to perform various tasks in a data processing pipeline. Each model has its strengths and weaknesses, and engineering a real-world application often involves solving many problems without a perfect answer. Your imperfect solution will nonetheless create a fun user experience!
-
Clone the repository and navigate to the downloaded folder.
git clone https://github.com/udacity/deep-learning-v2-pytorch.git cd deep-learning-v2-pytorch/project-dog-classification
NOTE: if you are using the Udacity workspace, you DO NOT need to re-download the datasets in steps 2 and 3 - they can be found in the /data
folder as noted within the workspace Jupyter notebook.
-
Download the dog dataset. Unzip the folder and place it in the repo, at location
path/to/dog-project/dogImages
. ThedogImages/
folder should contain 133 folders, each corresponding to a different dog breed. -
Download the human dataset. Unzip the folder and place it in the repo, at location
path/to/dog-project/lfw
. If you are using a Windows machine, you are encouraged to use 7zip to extract the folder. -
Make sure you have already installed the necessary Python packages according to the README in the program repository.
-
Open a terminal window and navigate to the project folder. Open the notebook and follow the instructions.
jupyter notebook dog_app.ipynb
NOTE: While some code has already been implemented to get you started, you will need to implement additional functionality to successfully answer all of the questions included in the notebook. Unless requested, do not modify code that has already been included.
NOTE: In the notebook, you will need to train CNNs in PyTorch. If your CNN is taking too long to train, feel free to pursue one of the options under the section Accelerating the Training Process below.
If your code is taking too long to run, you will need to either reduce the complexity of your chosen CNN architecture or switch to running your code on a GPU. If you'd like to use a GPU, you can spin up an instance of your own:
You can use Amazon Web Services to launch an EC2 GPU instance. (This costs money, but enrolled students should see a coupon code in their student resources
.)
Your project will be reviewed by a Udacity reviewer against the CNN project rubric. Review this rubric thoroughly and self-evaluate your project before submission. All criteria found in the rubric must meet specifications for you to pass.
Your submission should consist of the github link to your repository. Your repository should contain:
- The
dog_app.ipynb
file with fully functional code, all code cells executed and displaying output, and all questions answered. - An HTML or PDF export of the project notebook with the name
report.html
orreport.pdf
.
Please do NOT include any of the project data sets provided in the dogImages/
or lfw/
folders.
Click on the "Submit Project" button in the classroom and follow the instructions to submit!
- The submission includes all required, complete notebook files.
Question 1: Assess the Human Face Detector
- The submission returns the percentage of the first 100 images in the dog and human face datasets that include a detected, human face.
Use a pre-trained VGG16 Net to find the predicted class for a given image. Use this to complete a dog_detector function below that returns True if a dog is detected in an image (and False if not).
- The submission returns the percentage of the first 100 images in the dog and human face datasets that include a detected dog.
- Write three separate data loaders for the training, validation, and test datasets of dog images. These images should be pre-processed to be of the correct size.
- Answer describes how the images were pre-processed and/or augmented.
The submission specifies a CNN architecture.
Question 4: Outline the steps you took to get to your final CNN architecture and your reasoning at each step.
- Answer describes the reasoning behind the selection of layer types.
- Choose appropriate loss and optimization functions for this classification task. Train the model for a number of epochs and save the "best" result.
- The trained model attains at least 10% accuracy on the test set.
- The submission specifies a model architecture that uses part of a pre-trained model.
- The submission details why the chosen architecture is suitable for this classification task.
- Train your model for a number of epochs and save the result wth the lowest validation loss.
- Accuracy on the test set is 60% or greater.
- The submission includes a function that takes a file path to an image as input and returns the dog breed that is predicted by the CNN.
- The submission uses the CNN from the previous step to detect dog breed. The submission has different output for each detected image type (dog, human, other) and provides either predicted actual (or resembling) dog breed.
- The submission tests at least 6 images, including at least two human and two dog images.
- Submission provides at least three possible points of improvement for the classification algorithm.
(Presented in no particular order ...)
(1) AUGMENT THE TRAINING DATA Augmenting the training and/or validation set might help improve model performance.
(2) TURN YOUR ALGORITHM INTO A WEB APP Turn your code into a web app using Flask!
(3) OVERLAY DOG EARS ON DETECTED HUMAN HEADS Overlay a Snapchat-like filter with dog ears on detected human heads. You can determine where to place the ears through the use of the OpenCV face detector, which returns a bounding box for the face. If you would also like to overlay a dog nose filter, some nice tutorials for facial keypoints detection exist here.
(4) ADD FUNCTIONALITY FOR DOG MUTTS Currently, if a dog appears 51% German Shephard and 49% poodle, only the German Shephard breed is returned. The algorithm is currently guaranteed to fail for every mixed breed dog. Of course, if a dog is predicted as 99.5% Labrador, it is still worthwhile to round this to 100% and return a single breed; so, you will have to find a nice balance.
(5) EXPERIMENT WITH MULTIPLE DOG/HUMAN DETECTORS Perform a systematic evaluation of various methods for detecting humans and dogs in images. Provide improved methodology for the face_detector and dog_detector functions.
-
[] TRY BATCH NORMALIZATION (IN CNN FROM SCRATCH) https://pytorch.org/docs/master/nn.html#batchnorm1d
-
[] TRY ELU ACTIVATION FUNCTION (IN CNN FROM SCRATCH) https://pytorch.org/docs/stable/nn.html#non-linear-activations-weighted-sum-nonlinearity
-
[] TRY HIGHER DROPOUT RATES (40-50%)
-
[] REMOVE TRAINING IMAGES FROM TESTING IN THE LAST SECTION