AI & ML Projects

Detecting Lane Lines for Self-driving Cars

As part of Udacity Self-driving cars ND, I worked on developing a pipeline that detects lane lines from a video file. The pipeline consists of algorithms such as gray scale, gaussian blur, canny edge detection, and masking the area of interest. The model identifies lane lines by drawing a shaded red line.

The link below shows the pipeline on a Jupyter Notebook on my Github:

https://github.com/sami-aladawi/Lane_Lines_Detection_NDP1/blob/master/P1.ipynb

The output of the model is a video file with the lane lines detected.

Here is the output video of the model for white lane lines:

And here is the output video of the model for yellow lane lines:

Classifying Restaurant's Reviews

In this model, 1000 reviews of restaurants are provided as a data set. The goal is to train an NLP model to classify whether the review is positive or negative. Naive Bayes is used as the classifier but can achieve better accuracy by using other classifiers such as Maximum Entropy and CART.

The link to the Jupyter Notebook is below: