Many Filipinos living in urban areas rely heavily on roads and sidewalks to accomplish their daily routines. Given that only 31% of the nation’s households owns at least one car, quality public infrastructure must be made available and accessible to all people. However, this is not always the case. Hundreds if not thousands of sidewalks are either broken, narrow, unsafe, or uncomfortable to work with. With the use of our sidewalk rating and labeling platform, we hope to introduce a method of collecting sidewalk accessibility ratings in the Philippines.

Using your annotations as training data, we hope to train a machine learning model capable of assessing sidewalk assessibility. We use object detection and object segmentation models to assess the preliminary assessibility score of a sidewalk.

We are a group of Computer Science majors from the College of Computer Studies, De La Salle University, and we are currently working on building this sidewalk rating and labeling platform for our undergraduate thesis.