Getting Images
We first get the key frames from the camera situated in the upper right corner of the car. The camera is connected to the small computer which is the main component of the in-car system.
Getting Frames
Our Solution
MASK DETECTION
Mask Compliance Detection feature tracks and detects
the Masks on the people’s face in a given scene.
This feature tracks the people with or without masks in
a real time scenario and generates an alarm if there is a
violation in the scene. This feature helps in tracking
the people those who are violating the compliance of
wearing a mask in public places.
This feature requires a dedicated license to configure and use.
SOCIAL DISTANCING VIOLATION DETECTION
Social Distancing Violation detection feature traces and tracks the distance between two or more people with inbuilt analytics and algorithms. This feature tracks the two or more people in a real time scenario and generates an alarm if there is a violation in the scene. This feature helps in your premises to track & ensure social distancing is followed which in-turn a critical step for people to avoid being infected.
This feature requires a dedicated license to configure and use.
Our Model
Model: We decided to utilize a technique called transfer learning to make the learning faster and
better due to our limited dataset.
VGG16 We used VGG16 pre-trained network trained with 'imagenet'.
We changed the last few layers to adapt the network to our own dataset.
Dataset: Our model "learned" what distracted driving looks like from images obtained from a Kaggle Competition sponsored by State Farm Insurance. We performed histogram matching to try to adapt this dataset to the images we take by our own camera. The 10 classes of distracted driving we classify:
c0: texting with right hand
c1: texting with left hand
c2: talking to passanger
c3: talking on the phone with right hand
c4: talking on the phone with left hand
c5: safe driving
c6: reaching behind
c7: operating the radio
c8: eating/drinking
c9: doing hair/make-up
Accuracy: We achieved over %80 (84.7) testing with around 20k pictures with the dataset. However, as the dataset was a simulated environments and due to the outside factors, it does not generalize well for real-life images.
Dataset: Our model "learned" what distracted driving looks like from images obtained from a Kaggle Competition sponsored by State Farm Insurance. We performed histogram matching to try to adapt this dataset to the images we take by our own camera. The 10 classes of distracted driving we classify:
c0: texting with right hand
c1: texting with left hand
c2: talking to passanger
c3: talking on the phone with right hand
c4: talking on the phone with left hand
c5: safe driving
c6: reaching behind
c7: operating the radio
c8: eating/drinking
c9: doing hair/make-up
Accuracy: We achieved over %80 (84.7) testing with around 20k pictures with the dataset. However, as the dataset was a simulated environments and due to the outside factors, it does not generalize well for real-life images.
Prediction
Sending Data
We send the date, time, location, type and picture to the MySql database and the web server.
This data can be used for further researchs to minimize car accidents.
Sending Data
Warning
We warn the driver with an auditory alarm for 3 seconds.
User can select the alarm sound from web interface.
Warning
Giving Feedback
The users can view statistics and information about these distraction from our web application.
Location, time and frame of the distractions can be viewed so that users can check and notice
their fault.
View Distractions
How does it work?
Social Distancing Alert System AI platform uses existing IP cameras to identify if people are following social distancing. Social Distancing Computer Vision system finds the gap between two persons detected in the camera.
Then, you can freely enjoy life...
FEATURES AND BENEFITS
No new hardware required
You don’t need new cameras to enable the platform. Instead, it can work on the existing RTSP camera and connect to your existing smart speakers.
Different GUI Options
Easy to use interface with live monitoring and people count on the same interface
Real-time monitoring
to ensure staff and other patients are adequately informed and protected and if people within a zone are following social distancing guidelines
High Accuracy
Use of video analytics with high accuracy (> 90% with recommended operating conditions) to isolate and report non-compliance
Tech Stack
