Semantic Segmentation Project

Introduction

This project is based on the Fully Convolutional Networks for Semantic Segmentation paper from UC Berkeley. I replicated their model to detect whether each pixel is from the road or not.

Architecture

The encoder part of the FCN is a pre-trained VGG16 on ImageNet and the decoder is obtained by up-sampling the layers 3, 4 and 7 of that model. I applied regularization to each 1x1 convolutions as well as the deconvolutions.

Data

I used the Kitti Road Dataset to train the FCN. It is already split between training and testing sets.

Here is an example of image they provide and the associated labels:

Training Image	Labels

Hyperparameters

I trained the model on 2 Epochs with a batch size of 2 to obtain the following results. The dropout probability is 0.5. The learning rate was set to 1e-3. Regularization was applied with beta = 1e-3.

Results

Road Classified Images

---

Original README from Udacity

Setup

GPU

main.py will check to make sure you are using GPU - if you don't have a GPU on your system, you can use AWS or another cloud computing platform.

Frameworks and Packages

Make sure you have the following is installed:

• Python 3
• TensorFlow
• NumPy
• SciPy

Dataset

Download the Kitti Road dataset from here. Extract the dataset in the data folder. This will create the folder data_road with all the training a test images.

Start

Implement

Implement the code in the main.py module indicated by the "TODO" comments. The comments indicated with "OPTIONAL" tag are not required to complete.

Run

Run the following command to run the project:

python main.py

Note If running this in Jupyter Notebook system messages, such as those regarding test status, may appear in the terminal rather than the notebook.

Submission

1. Ensure you've passed all the unit tests.
2. Ensure you pass all points on the rubric.
3. Submit the following in a zip file.

• helper.py
• main.py
• project_tests.py
• Newest inference images from runs folder (all images from the most recent run)

Tips

• The link for the frozen VGG16 model is hardcoded into helper.py. The model can be found here
• The model is not vanilla VGG16, but a fully convolutional version, which already contains the 1x1 convolutions to replace the fully connected layers. Please see this forum post for more information. A summary of additional points, follow.
• The original FCN-8s was trained in stages. The authors later uploaded a version that was trained all at once to their GitHub repo. The version in the GitHub repo has one important difference: The outputs of pooling layers 3 and 4 are scaled before they are fed into the 1x1 convolutions. As a result, some students have found that the model learns much better with the scaling layers included. The model may not converge substantially faster, but may reach a higher IoU and accuracy.
• When adding l2-regularization, setting a regularizer in the arguments of the tf.layers is not enough. Regularization loss terms must be manually added to your loss function. otherwise regularization is not implemented.

Using GitHub and Creating Effective READMEs

If you are unfamiliar with GitHub , Udacity has a brief GitHub tutorial to get you started. Udacity also provides a more detailed free course on git and GitHub.

To learn about REAMDE files and Markdown, Udacity provides a free course on READMEs, as well.

GitHub also provides a tutorial about creating Markdown files.