Tesla FSD Beta V9 Visualization Is a 'true mind’s eye of the neural net,' Says Elon Musk

by Eva Fox May 01, 2021

Tesla FSD Beta V9 Visualization Is a 'true mind’s eye of the neural net,' Says Elon Musk

Full Self-Driving (FSD) Beta, introduced by Tesla last year, continues to improve, offering users an even wider range of options, and ultimately leading to Level 5 autonomy. In addition to expanding its capabilities, Version 9.0 will have improved visualization, which should instill more peace and confidence in drivers.

For some time now, Tesla has been striving to move neural networks (NN) to surround video, which will allow cars to reach the next level of FSD. According to earlier comments by Tesla CEO Elon Musk, it seems that the company has made progress in this, and the next version of FSD Beta V9.0 can work by receiving information only from cameras, removing radar from vehicles. However, this is not the only improvement.

Along with a huge leap forward in achieving full driving autonomy, the FSD Beta V9.0 will have significantly improved visualization. At the moment, the objects surrounding the car are depicted in the form of parallelepipeds, cubes, and dots. Thus, drivers have the impression that they are in developer mode, which can both add positive emotions and influence drivers, causing increased anxiety. Of course, Tesla does not plan to leave the visualization in this form, so the company is working on improving it.

Current rendering



According to Musk, the visualization of FSD Beta V9.0 is "a real mind's eye of a neural network." @WholeMarsBlog/Twitter suggests that this means a smoother rendering, and instead of the previous representation of the uncertainty, drivers will be able to see a visualization of a probability distribution.


Musk also wrote that "FSD display V9.0 will show actual probability distribution of objects - true mind's eye of the neural net. This is so cool!" Probability distributions are theoretical distributions based on assumptions about a source population. The distributions assign a probability to the event that a random variable has a specific, discrete value, or falls within a specified range of continuous values. Probability distribution objects allow to fit a probability distribution to sample data, or define a distribution by specifying parameter values. Thus, we are talking about the fact that the visualization in the new version of FSD Beta will be smoother and more familiar, and at the same time, it will not contain unnecessary information that is not needed by Tesla drivers.

© 2021, Eva Fox. All rights reserved.

_____________________________

We appreciate your readership! Please share your thoughts in the comment section below.

Article edited by @SmokeyShorts, you can follow him on Twitter





Previous  / Next