Due to inadequate understanding of these road dangers, drivers and other road users frequently encounter risks.Speed breakers and potholes pose significant threats to road safety, vehicle integrity, and passenger comfort. Without timely detection and alerts, drivers are often caught off guard, leading to accidents, vehicle damage, and potential injuries. Therefore, there is need to develop reliable and efficient speed breaker and pothole detection systems that can address these challenges and enhance overall road safety and infrastructure maintenance.
The main challenge was to procure proper required data from open source to run the deep learning model. We also faced a major challenge whie pre-processing the dataset to the required format. Since we were using a online mode of GPU for training the models, it had certain time and space limit due to which we had to switch google accounts now and then.