![]()
Visiting guest. Why not sign in?
|
|
Reply to Message
Not registered yet?
The AI Depot has a focused community of friendly users. Rather than let anyone abuse the site at the brink of promiscuity, we prefer to let only those with an active interest participate... this simply requires registering.
Why not sign up!
Joining the site's community is completely free. You can then post messages freely, and customise your personal profile at will. Specific privileges will also be granted to you, like being able to access printer-friendly articles without restrictions. So, why not register?
Parent Message
Attempts to Supercede Human Vision
UCLA researchers are attempting to model and extend human vision, and implant it into working robots in order to detect specific visual patterns. The article over on Science Daily entitled Computers With Human-Like Vision Could Strengthen Security And Surveillance, goes into this in more detail. Essentially, the researchers claim that to extract complex distance/angle and other 3D information from 2D images, they need a time based sequential analysis of these flat images. This bothers me. Indeed, instead of processing numerous 2D images to provide 3D data, human vision is based on stereoscopy. Analysing two 2D images at each point in time provides most of the essential 3D information. Admittedly, things like shadow and lighting changes over time can also provide this information in a more complex way. None the less, this still seems overly complex. Why not use sonar or laser sensors to get this distance information? Understandably, such sensors would be quite expensive when used in large quantities, and would become more imprecise at the distance increases - or when the weather conditions are bad... but this is fundamentally easier and faster than having to deal with 2D feature extraction algorithms. Regardless, the ways of obtaining this 3D information about the scene should be treated as a black-box, so that the AI can process it afterwards regardless of how it was obtained. |
|
Visual Extraction of Detail
In follow-up to the initial article, Info Space has posted an article that explains how humans can extract distance information based on angular declination. The article is called Brain Finding May Finetune Virtual Reality. An interesting confirmation of the motivation for the afore mentioned reseach, if nothing else! |
|
Back to the Artificial Intelligence Depot.