AI News, Using neural networks to detect car crashes in dashcam footage

Using neural networks to detect car crashes in dashcam footage

That’s why when you’re asked to prove yourself human through those CAPTCHA tests, you’re always asked a ridiculously simple question, e.g., whether an image contains a road sign or not, or selecting a subset of images that contain food (see Moravec’s Paradox).

In this post, I will describe how, as a Fellow for Insight Data Science, I built a classification machine learning algorithm (Crash Catcher!) that employs a hierarchical recurrent neural network to isolate specific, relevant content from millions of hours of video.

For businesses that may have millions of hours of video to sift through (for instance, an auto insurance company), the tool I created is extremely useful to automatically extract important and relevant content.

In the ideal world, the perfect data set for this particular problem would be a large repository of videos with thousands upon thousands of examples of both car crashes and non-crashes.

Each video would also contain clear metadata and be consistent in terms of video (e.g., location of camera in car, video quality and duration) and content (e.g., type of car crash, whether it be “head-on” or “t-bone”).

Their videos were short, four second-long clips with approximately 600 videos of accidents and about a thousand videos without accidents (just normal, boring driving scenes).

If I were to train an algorithm on the 439 negative examples and 36 positive examples in my data set, the resulting model could easily predict that there were no crashes and still have 92.5% accuracy.

Each image can be thought of as a two-dimensional array of pixels (originally with a size of 1280x720), where each pixel has information about the red, green, and blue (RGB) color levels which creates a 3-D shape.

Video datasets are particularly challenging because of their structure — while each frame in the video might be understood using standard image recognition models, understanding overall context is more difficult.

This approach allowed me to train a model to understand the flow of features and objects within a single video, and translate that into patterns that differentiate videos with crashes from videos without crashes.

The first neural network analyzes the time-dependent sequence of the images within each video, tracing objects or features as they move or change throughout the clip (e.g., car headlights or car bumpers).

The second recurrent neural network takes the patterns and features encoded by the first neural network and learns patterns to discern which videos contain accidents and which do not.

I tuned the model’s hyperparameters (e.g., number of layers of neurons, number of videos loaded into memory at a time, loss function, number of epochs) to optimize for accuracy, where the training and validation sets iterated through different options of these hyperparameters.

You may be wondering about that other 20% of the data I haven’t mentioned — these 52 videos were my holdout test set to analyze the final performance of the model after tuning the model.

Reducing false negatives (crashes incorrectly identified as normal driving scenes) is key, as these are the cases that are most important to companies sifting through large quantities of video, even if there are occasional videos without crashes predicted to have crashes (false positives).

To train this algorithm to accurately predict a wider variety of situations (e.g., classifying head-on collision vs rear-end collision vs car-truck collisions), more data is necessary.

Crash test measurement

Automated car body scanning supported by 3D Arena.

Inside IIHS: Crash test photography

The "Inside IIHS" series of videos offers a behind-the-scenes glimpse of the work of the Insurance Institute for Highway Safety's Vehicle Research Center. "Inside ...

Aerial Photogrammetry for Traffic Crash Investigation

2019 Lamborghini Terzo Millennio - Self-Healing!!

Lamborghini Unveils a Self-Healing, Electric Supercar The Terzo Millennio concept is what can happen when MIT gets in the car game. The next generation of ...

SHARK SONG on RAFT! Survival Game w/ Baby Shawn in Danger! 1st Night Minecraft? FGTEEV Gameplay/Skit

Poor Duddy gets bothered by a Shark Attack while singing his Shark Song on a Raft he built in the middle of the ocean! We are trying to survive the 1st night like ...

DUAL CAMERA CAR DASH CAM - 2.7" 1080P Full HD DVR - Dome G30B

Gear Best Link for this Dash Camera: Gear Best Car DVR Category: ..

Pre- and post-crash measurement with AICON MoveInspect XR8

Based on the renowned AICON MoveInspect Technology, the XR8 is the only technology able to simultaneously capture multiple points. This makes the system ...

Why Don't Planes Crash Into Each Other More Often?

With all the planes in the sky at once, what kinds of technology are used to prevent mid-air collisions? Squarespace link: Visit ..

How to fix problems with the Samsung Galaxy S7 Edge

When your Samsung Galaxy S7 Edge is having issues, there are a number of ways to solve them. Here is a collection of common problems with the Galaxy S7 ...