Skip to content

The reason Tesla FSD performs better with just computer vision than Lidar

twitter.com
4 pointslopkeny12ko2 comments
On HN

Comments

We know it is sufficient because all humans drive using just vision.

Wrong!

Humans use vision, sound, feel/touch, knowledge of environmental conditions that can't be seen (cross wind, black ice) and can judge the likely response of other human drivers to unexpected and unpredictable road conditions.

If a human driver sees an accident or perhaps a tire on a large truck disintegrate up ahead on the freeway, he immediately knows to change lanes in order to avoid what is coming his way.

Can camera only FSD do this?

Disclaimer: wholemarsblog is sponsored by Tesla.

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.