Bestand wählen

Unity Game Engine

Zitierlink des Filmsegments
Embed Code

Automatisierte Medienanalyse

Erkannte Entitäten
what do we do we want to and brute blue blue 2 blue wanting to fund and some
things I would change emission you guys saying platforms so you need to imagine how we create a physic physical simulations and the game this is a graphics and decreases simulation of some kind some kind of simulation virtual environment and I will show you today a as a the identity that along the lines indicated data coming at unity and interview you anomaly detection on virtual environments so here's a this is that is mostly about how to get along with demo of getting the data out in Internet and then getting anomalies out of nutrients into it so I haven't binding to the knowledge models really so that's again this kind of natural thing coming out of works so the way it works is there's a so you can have a growing in Python and stream data out of unity interval into server which think reads from and and then it's in the background so you understand the client which is a new pair which is given model that can get in return explorers and then
here's a simulation of music cars driving around and what you do is on its going to send its global positioning like GPS position in this virtual world using the geospatial do anomaly-detection on its GPS coordinates over time this is
a list of all have a so as it is done until you in underwriting driving and you see that the traces behind right so is this is not yet as a new patent everything that novelists and designate this couple times of the solution to see something that you need to be thinking about the predictions as well as patterns and knowledge sources of them and the use of couple times eventually everything becomes pretty which creates a knowledge of the positions of a slightly different material in the exact same thing Time and the ability to generalize this to use vision and certainly not going happen on and when l I think it's something about the so called greenhouse
because when we look at the world and just like to have always becomes red and it's detecting anomalous behavior and so and so on and go back and to look at the the recognized the analysis and and I can't even during to the opposite direction temporal sequence of bits of detecting that some of the novels and the effects less knowledge lectures so basically this
GPS is like the sensor on the spot on GPS sensor that is sending its students you this by 1
of proxy server and you think is basically making requests that he was the 1st to really get the GPS coordinates these initial model returns the anomaly score to the processor which sent you so there is this what is going on here
but this gives repositories and there's instructions on how to break yeah i it is important point in and then use certain along lines of code and you use the application to send data you begin reading about you and then you just on the posterior you get the so here's the deal with the package the packaging important and then this all this because of because this them owing to show you is here this is the real long and can use that as the starting now starting point for if you want to have you on this is that in the model so yeah that's available there yes that along the path from yes so this that's a question about the geospatial knowledge and so on and so the question is can learn multiple patterns of multiple like rabbits and and yes a cancer you in a good program horizontally across a set of this of vertical loop can recognize that has had its anything outside of those 2 would be anomalous so it never went on and on and this money if you have more questions about it'll is really a geospatial knowledge action and previously was also the because that is also online think in in under the new mentioned relative indifference so you check out as well so if you want to the thing the we have tell you every passing the limit of what it's very in in the temporal memory but it's very high and so it's not really an 3 old stuff that easily found it was you know it's really it's also without the signal from the noise so and that's why it takes some time to learn a sequence of your models and so they can find inconsistencies in the on light of forgetting that is strongly learning takes a long more time and learn which is going to give a much multiple times right so you you could type it when you could have multiple ones and at the interface for this year the API that property is just it doesn't really care what the inputs and outputs are you just you can define them in Europe in the project so you can have multiple coordinates coming off of multiple cars going to use the word into the bottom side and you have multiple different models anomaly-detection 2012 the use of yes any other questions but also this goes here and out of the users interested in you have been it is because there's a thing of the how do you have to so so my constant is that there is a lot of work on the mind
Demo <Programm>
Virtuelle Realität
Natürliche Zahl
Streaming <Kommunikationstechnik>
Gruppe <Mathematik>
Diskrete Simulation
Folge <Mathematik>
Temporale Logik
Maschinelles Sehen
Proxy Server
Folge <Mathematik>
Kartesische Koordinaten
Inverser Limes
Kategorie <Mathematik>
Temporale Logik
Wort <Informatik>
Projektive Ebene


Formale Metadaten

Titel Unity Game Engine
Serientitel 2015 Spring NuPIC Hackathon
Anzahl der Teile 19
Autor Sample, Hack
Surpur, Chetan
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 10.5446/18066
Herausgeber Numenta Platform for Intelligent Computing (NuPIC)
Erscheinungsjahr 2015
Sprache Englisch

Inhaltliche Metadaten

Fachgebiet Informatik
Abstract Chetan Surpur demonstrates a project he created that does anomaly detection on coordinates streaming out of the Unity Game Engine.

Ähnliche Filme