3 ms·
The camera portion of Tesla's vision system produces a grayscale image that's then run through some algorithmic pattern detection to figure out what's what in t
by niftich 10y ago
The camera portion of Tesla's vision system produces a grayscale image that's then run through some algorithmic pattern detection to figure out what's what in the scene.
Here is an actual dataset [1] of stills from Daimler that's used to train algorithms for pedestrian detection.
The workings of this system are similar to the cameras that are mounted above intersections to detect waiting cars (instead of sensors embedded into the roadway) [2]
[1] http://www.gavrila.net/Datasets/Daimler_Pedestrian_Benchmark_D/Daimler_Mono_Ped__Detection_Be/daimler_mono_ped__detection_be.html http://www.gavrila.net/Datasets/Daimler_Pedestrian_Benchmark...
[2] http://www.modot.org/stlouis/links/signalcameras.htm http://www.modot.org/stlouis/links/signalcameras.htm
- mkj 10y agoSurprising it doesn't make use of colour, it'd be throwing away quite a bit of differentiating information?
- bobsil1 10y agoVery common in image ML, for speed. Results usually good enough.
- Neeek 10y agoI work with imaging systems that identify medicines. While they are wildly simpler in terms of analysis (basically using the silhouette to calculate shape and size), most attempts to add colour to the equation ends up adding a new, complex element that the inspection can fail on. Mostly because colour is inconsistent under varying lighting conditions. Again, very different systems, but might be relevant when considering if straight contrast is enough to work with or not.