A Whole New World [DEMO #4]

Zitierlink des Filmsegments
Embed Code

Automatisierte Medienanalyse

Erkannte Entitäten
the area of a reader of upset some text test statistics because the it in the US in this book that he is in the middle of the building and the user's utilities and in so I'll show you a little robots based on so this is actually the 1st the 1st experiment do you want think and has and this the set is very similar to how small around here and I ask this is taking the standard features of objects actually each of us find it will have to pay for it so that it also had a motor on the back of the head of the department that you have high quality programming to randomly sampled from the start of the considerable distances people and that and then you can get some information about the layout of the world and if you also if you change the spatial configuration of the 1 that I can say think of this is a new which of to act as a little bit of theory and we show you that you can understand what's going on what we're doing is taking taking this data from the sensor node in the in the history of the universe the business of both of them can easily for and so if you think about it from the perspective of the role of what it's saying is it you think of overlap now and then it moves sensors of away the object using inside our senses and then you can just write it out and turn right now what I expect to see how far we largely is supposed to be and to make a prediction so in this case where for its information about the current sensor readings of our logic looking at is and motor commands that execute and later for for basically where those sensorimotor transitions and then to predict what's going to scene with the center value developed so if this was predicted by the for of the transition is pretty by the for successfully into another vision and now there's 2 3 control over the predicted transition because now we want more predictable we can build a stable representation that well and selectively pulls over at properly and elaborated in this new suppressing the form and passed through a chain so we want to see is sort of a stable representation once the world comes so on and they can be used as part of the reason for that so you look at the end of this paper is the Top provides the user very likely belong to the work of the yesterday in the form other words that the teacher and that's 3 objects this is all the sensory
device and since these objects and if moron this sample the outer movements on and extractors movements around the center of the ways that it GATExplorer transitions as seen on the client side of exploring some of its initial time so users while there because although it is not in the back of it so it just have to about 30 seconds 30 random movement this is by the way way that nothing in training the models and the the eliminating at this point so here
we're using here is the a
representation of the form that all of these things that really fast and the number of predictor cells in the war that was that was initially had going on the predicted that so many combinations of the bond that there is problem this can happen and that the classifier successfully this world view the stabilization of the region showing that that's all that's called the best interests of the universe so users and the ability the now in
there in this world they were the thinking of of world In the world of
the bold part of the of the government and this imagine just change the speed of the reaction so should be the mind the mind so that we can go back to that the 1st world but the are variables and with different was the ways on what the what was the involvement of the state of the that's what you have written answers think of where 3 this is no stability is changing the observations are and this is a start making predictions were therefore the United States and the world the not head of the focus of representation here is stable right so but it's different from this is that the real world and the faster the years these things here that we will not the ones and it also plays the sound and how do you go back to the 1st world right it can become the features the which catalog the nation's there is some small variations and that always the sensory motors and an arm movements here but hopefully it can be passed the then you have to think about how much the only thing that we can know that only 1 the world and you know even tho it was on the same side of the picture which includes not the fact that you have in the noise is this is the sum of the the year of because of what is going on and on and on the other hand is this is is not the end of of the of the of the of the
of the of the of the of the of the users of these some of them on the exam is more than of the things that know yeah the documents from the of it and you will be the same word
again but this time you want but what do you this paper or book and who was this this new reality so that's something think that somebody you can see but also some other british somebody because you see there's no stable representation theory of bias in word want here but it has a lot of the know times and those of all of the new problem was that it was many of the
problems that we have but know the future knowledge of world I was consider the release this is that this is 1 of things that you or by the way with the of the of the of the of the image in the in the probability of the data at the time and I'm going on and on and on and on and on the I of
Mathematische Logik
Physikalische Theorie
Demoszene <Programmierung>
Maschinelles Sehen
Gebäude <Mathematik>
Objekt <Kategorie>
Verkettung <Informatik>
Wort <Informatik>
Lesen <Datenverarbeitung>
Objekt <Kategorie>
Innerer Punkt
Stabilitätstheorie <Logik>
Zellularer Automat
Stabilitätstheorie <Logik>
Gewichtete Summe
Rechter Winkel
Hydrostatischer Antrieb
Wort <Informatik>
Stabilitätstheorie <Logik>
Lemma <Logik>
Lineare Darstellung
Wort <Informatik>
Bildgebendes Verfahren
Demo <Programm>


Formale Metadaten

Titel A Whole New World [DEMO #4]
Serientitel 2014 Fall NuPIC Hackathon
Anzahl der Teile 19
Autor Surpur, Chetan
Cui, Yuiwei
Lizenz CC-Namensnennung 3.0 Unported:
Sie dürfen das Werk bzw. den Inhalt zu jedem legalen Zweck nutzen, verändern und in unveränderter oder veränderter Form vervielfältigen, verbreiten und öffentlich zugänglich machen, sofern Sie den Namen des Autors/Rechteinhabers in der von ihm festgelegten Weise nennen.
DOI 10.5446/18068
Herausgeber Numenta Platform for Intelligent Computing (NuPIC)
Erscheinungsjahr 2014
Sprache Englisch

Inhaltliche Metadaten

Fachgebiet Informatik
Abstract A demonstration of sensorimotor inference in simple robotics.

Ähnliche Filme