Because such algos AFAIK only work on v. similar images/objects with limited noise.
If objects are reasonably similar in form, such algos can effectively average them – or perform similar operations. And you can *see* how these algos work – i.e. what the similarities of objects presented, are. But these objects/chairs are not at all similar in form, and you’re not in any way explaining what their similarities/commonalities are, or how the algos will find them - just praying that such similarities exist. There’s a book by De Bono called Practical Thinking which is relevant here – as to vague thinking. He presented a class with a black cylinder on a desk, wh. after a few mins. fell over. He asked them to explain what had happened. People, he pointed out, came up with many degrees of explanation - from very vague “ something (or “a mechanism”) inside made it fall over” all the way to fully detailed diagrams of what the mechanism was and how it made the cylinder fall over. Your (and most AGI-ers) thinking is basically more towards the vague “mechanism inside”/”algo” end (with one qualification – see below). You have to come up with a definite idea as to how the cylinder falls over – in this case what the visually similar elements are – and/or by what operations initially dissimilar elements can be shown to be similar. (Todor has made a vague stab at the latter – but hasn’t visually demonstrated it in any way). But you are basically avoiding the problem – just as all AGI-ers are basically avoiding any direct confrontation with the unsolved problem[s] of AGI – of wh. this is one manifestation. Here comes the qualification – AGI-ers do go into great detail about their architectures, logics etc - wh. to them sounds technically impressive – but they never show how these architectures or logics apply/ are relevant to the actual problem. Similarly some kids in the class came up with very precise descriptions of some mechanism or motor inside the cylinder – but didn’t explain how it made the cylinder fall over. From: Piaget Modeler Sent: Sunday, November 04, 2012 10:36 PM To: AGI Subject: RE: [agi] The Fundamental Misunderstanding in AGI [was Superficiality] Here I go again, back into the vortex... If the task is recognizing chairs, why wouldn't a simple constructive induction algorithm work. Run it on the first N images and test on the remainder? For each test image also present the category "Chair" to the process. Why wouldn't this successfully classify the chairs or chair portions in the images you presented? -------------------------------------------------------------------------------- From: [email protected] To: [email protected] Subject: Re: [agi] The Fundamental Misunderstanding in AGI [was Superficiality] Date: Sun, 4 Nov 2012 22:06:56 +0000 Well, Todor, these are pictures of chairs and *you* have little problem recognising them – (agreed, the odd one may cause confusion – but that is a fundamental part of the business of object recognition). And you can recognise sensory images of such chairs from a distance, with v. limited “3d/solid” information (it sounds like you’re talking mainly about viewing objects as “solids” close up (in the flesh). We detect a lot of “3d” info, you see,. from 2d pictures. If you are suggesting that such object recognition processes in humans and animals are based on “solid” experience of comparable objects, I would agree. But then you will have to give some indication of how the methods you outline apply to robots interacting with solid objects – I wasn’t aware that you were/are doing that. Are you? Ultimately you will have much the same problems dealing with solid objects as with their pictorial images. They too are as formally diverse as the images. Then certainly you/a robot can interact physically with the objects and ascertain if you can sit on them. But there will still be problems of classification – for example, separating stools/walls/boxes and other sittable objects from chairs. And more or less each solid chair in the pictured range still represents an extraordinary **transformation** of the last, You’re not really advancing our, i.e. the general, analysis of the problem, by implying it all can be easily solved. – by geometric operations. You have provided no evidence that they apply to those images, and by extension to their objects – and can explain the uniformities underlying that vast range of transformations. I contend as you know, that it is only by comparing them with the multiform transformations of irregular, more or less fluid objects ranging from blobs and waterdrops all the way to rocks and wood chunks, that we can understand the design transformations of chairs - and not by comparing them with the geometric, uniform transformations of geometric objects. I would suggest it is far better for you to recognise the difficulties of the problem – which you have effectively done by pointing out the enormous visual ambiguities of the chair images – with wh. I broadly agree. Then, I suggest, you have more to offer – it’s useful for example to bring in, as you have done, how the recognition of objects depends on how we physically interact with those objects – whether in this instance we can sit on them. (This exchange with you & Aaron has been valuable for me in underlining how probably essential that dimension is for object recognition). But unless I missed it, your proposals don’t call for simulated interaction of the robot’s/viewer’s body with the objects viewed – such simulation is certainly necessary for us to recognise those chairs- and there is massive evidence for it. (Is anybody in AGI or robotics attempting such simulation?) . AGI | Archives | Modify Your Subscription ------------------------------------------- AGI Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/21088071-c97d2393 Modify Your Subscription: https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-2484a968 Powered by Listbox: http://www.listbox.com
