Safety Assistance for Cars

• 

  wall in front

• 

  speed bump ahead when image shown to AI

• 

  Shows result for a human in front of camera(simulates human in front of car)

Inspiration:

Many people may get distracted while driving which causes 1000s of accidents. So we have decided to help them using AI.

What it does:

Using image classification we made it so that it will notify the driver if there is a wall, pedestrian, or a speed bump in front of the car in case the driver was not paying attention.

How we built it:

Using deep learning, keras.module, we have made an AI with a large amount of data compiled using Google Teachable machine to classify images in real-time to detect whether there is an obstruction or a potential threat in front of the car.

Challenges we ran into:

Since it was our first Hackathon we faced difficulties in coming up with a good idea for the people within 20 hours.

Accomplishments that we're proud of: We have managed to make a working AI within twenty hours without using the help of AI and only our knowledge.

What we learned: How we can implement AI in our daily lives to help others using Keras and TensorFlow.

;

What's Next for Safety Assistance for Cars: We plan to add more classes to the dataset and plan to improve the data quality within the dataset. We plan to implement this in many cars as it is very useful for drivers soon.

Built With

• keras
• opencv
• pygame
• python
• pyttsx3
• tensorflow