RSS
 

Arquivo para a ‘Visão artificial’ Categoria

Philosophy, Artificial Intelligence and God

24 Nov


We continue reading Peter Kreeft, now the interlocutor of Sócrates,dedodedeusia Peter Pragma is urgently to solve his problem of profession and seeks Socrates.

Peter started his philosophy classes and the teacher asked the question: what is the difference between human intelligence and computer intelligence, this is what the idea of going into computing referred to, we are on page 55, to which Socrates asked:

SOCRATES: You want to become a computer program, you mean? For the only thing that goes into computers. It’s the only language computers understand.

The teacher has thrown this question that Socrates considers it easy to answer, but Peter does not want any evasion, and Socrates goes straight to the point:
SOCRATES: Artificial intelligence can not do what your natural intelligence has just done.

 PETER: What?

SOCRATES: She keeps doing it.

PETER: Oh … oh. Asking questions, that’s it.

SOCRATES: Congratulations. You find the treasure.

Peter and Socrates are talking and realize that the very origin of sophisticated programs, such as artificial intelligence, should be spoken of in “people” and not just in machines.
But what is the beginning of all this? First we need an unscheduled programmer, or a programmer who can question your schedule, there is an initial piece of dominoes.
PETER: This sounds like a new argument for the existence of God.

SOCRATES: The same principle works in one case as in the other. It is the principle of causality, which says that you can not give what you do not have, that effects can not exist without proper causes, that there can be no less in the total cause than in the effect.

PETER: I do not know about God. Let’s talk about something we know: ourselves.

SOCRATES: Something you know, maybe. As for me, I find myself a mystery, just as I find God a mystery.

Peter will remember Socrates’ phrase: “Know thyself,” and say,
PETER: Why is it so difficult to know yourself, Socrates?

SOCRATES: Because the self is the knower himself. How can the subject become his own object? How can I become this? That’s why I also think God is a mystery. The human self is an image of the divine Self.

 And continue …

 

Pokémon Go and augmented reality

09 Aug

I arrive at my university to work and see a lot of students walking augmentedRealityalone or in groups there here, I thought it is-new strike, some proof of selection for a course, but they were what are now being called pokeballs capturing Pokemons but the scene was new and curious.
Since the past decade’s work and augmented reality environments, I thought a more productive use, virtual museums, studies of physics, chemistry or astrophysics; but the games were the first major use, it has become a fever and fever.
He told me a student who is not only out and go capturing Pokémon, incubating eggs, there are several tactics and mechanics that can become a better player than the others.
He summarized that way: not disperdice pokeballs, Pokemons with yellow circles around are very strong, look different Pokemons in different environments, capture Pokemons repeated that it is easier, and need not be connected with the camera the whole time.
But back to augmented reality, the Brazilian Congress in the area have been around since the end of the last millennium and the pioneer Dr. P. Milgram explains that actually are 4 levels: augmented reality, augmented virtuality and virtual environment, such as the Pokemons are the first level we think that there will be at least 3 more new developments.
What comes next may be even more surprising, then I think it is here to stay.

 

 

Augmented Reality and Pokemon Go I arrive at my university to work and see a lot of students walking alone or in groups there here, I thought it is-new strike, some proof of selection for a course, but they were what are now being called pokeballs capturing Pokemons but the scene was new and curious.
Since the past decade’s work and augmented reality environments, I thought a more productive use, virtual museums, studies of physics, chemistry or astrophysics; but the games were the first major use, it has become a fever and fever.
He told me a student who is not only out and go capturing Pokémon, incubating eggs, there are several tactics and mechanics that can become a better player than the others.
He summarized that way: not disperdice pokeballs, Pokemons with yellow circles around are very strong, look different Pokemons in different environments, capture Pokemons repeated that it is easier, and need not be connected with the camera the whole time.
But back to augmented reality, the Brazilian Congress in the area have been around since the end of the last millennium and the pioneer Dr. P. Milgram explains that actually are 4 levels: augmented reality, augmented virtuality and virtual environment, such as the Pokemons are the first level we think that there will be at least 3 more new developments.
What comes next may be even more surprising, then I think it is here to stay.

 

(Português) Óculos de VR chegaram

01 Mar

Sorry, this entry is only available in Brazilian Portuguese.

 

Rede transforma texto datilografado em manuscrito

17 Sep

Um novo programa de computador desenvolvido Manuscritopela Universidade Alex Graves de Toronto de desenvolveu um programa usando redes neurais recorrentes (RNNs) para converter textos digitados em tipo de escrita “orgânicos” isto é, praticamente reais.
O programa permite que usuário configure para moldar tanto legibilidade como o estilo de manuscrito que seja parecido ao de determinada pessoa ou a ser tipo de caligrafia, conforme artigo da revista Techlÿ.

O programa de Toronto embora gere longas RNNs de memórias similares de curto prazo combinando com pontos de dados individuais em sequências complexas baseadas em previsões de probabilidade.

Muitas redes neurais que dependem de aprendizagem, nome dado para o “treinamento” das redes neurais, para produzirem os mesmos padrões de resultados depois de uma escolha ou seqüência de escolhas, que torne o processo de aprendizagem como bem-sucedido.

Embora seja raro que as RNNs gerem uma mesma solução mais do que uma vez, pois seriam necessário uma grande padronizada de dados tanto sistema complexos de entrada como na saída de dados complexa e deveriam produzir resultados ponto-por-ponto.

Isto indique que o modelo de Toronto são RNNs ssuperiores a outros modelos de inteligência artificial, como por exemplo é o caso de sistema de RNN é mais capazes de modelagem de dados multivariados.

Isto porque as amostras RNN de sua própria saída, alimentação cada ponto de dados de volta para o algoritmo como uma nova entrada, os resultados são bem próximos a escritas reais.

O artigo já está submetido ao repositório ArXiv.

 

Poderá a internet ser emocional ?

01 Sep

Segundo o site TechRunch isto não é um conceito Emotionteórico, mas já há esforços para medir e quantificar a emoção humana de forma que a máquina possa compreender.

Três empresas já trabalham nisto, a Affectiva, uma empresa de medição de tecnologia de emoção, que coletou mais de um milhão de quadros com expressões faciais espontâneas, o VIBE, um software de visualização e interação da divisão de entretenimento da Microsoft Research, que emprega sensores para discernir humores via Bluetooth, e a Spire, que fabrica um pequeno sensores em forma de pedra que pode ser colocado em algum acessório do corpo humano, que pode medir o stress.

A maioria destas investigações e desenvolvimento são parte dos campos emergentes da computação afetiva, que examina maneiras de criar sensores e computadores que podem detectar e responder a emoções humanas.

Os estudos investigam padrões de fala, expressões faciais, gestos corporais e reações fisiológicas a estímulos específicos, mas será necessário acumular uma grande base de dados de emoções que pode treinar computadores para reconhecer e interagir de acordo com estímulos, uma vez que as pessoas podem reagir de modo diferente a um estímulo.

 

BBC develops remote control with the mind

29 Jun

Em parceria com a empresa This Place, a divisão de projetos digitais daiPlayer BBC desenvolveu em conexão com os conteúdos uma plataforma chamada iPlayer, a partir da qual pode-se quais programas assistir por meio de um aparelho que lê ondas cerebrais do espectador. Segundo Cyrus Saihan, chefe de desenvolvimento de negócios da divisão digital, em entrevista divulgada no site da BBC, o primeiro teste foi feito com dez funcionários da BBC e conseguiram usar e acionar o iPlayer, iniciando a transmissãod e um programa da TV. Saihan explicou, conforme o site, que este tipo de tecnologia poderá ser usada por pessoas com tenham algum tipo de deficiência que impeçam de usar o controle remoto. Ainda segundo ele, o o protótipo foi “feito para dar a quem faz nossa programação, nossos especialistas em tecnologia e outros usuários uma ideia de como isso poderá ser usado no futuro”, disse em entrevista ao site.
In partnership with the company this place, the digital arm of BBC projects developed in connection with the contents of a platform called iPlayer, from which one can watch what programs through a device that reads brain waves of the spectator.
According to Cyrus Saihan, head of development of the digital divide business, in an interview published on the BBC website, the first test was made with ten employees of the BBC and were able to use and press the iPlayer, starting transmissãod and a TV program.
Saihan explained, according to the site, that this type of technology could be used by people with have a disability that prevents them from using the remote control.
Also according to him, the the prototype was “made to give those who make our programming, our technology experts and other users an idea of how it can be used in the future,” he said in its interview.

 

Everyday scenes in 4D

24 Feb

Professor of Computer Science Noah Snavely, from Cornell University and his student Kevin Matzen have added a fourth dimension5Pointz of time to your system called Scene Chronology, and have collected millions of images from photo sharing sites on the Internet and views from popular sites taken from several different angles created from 3-D models combined, and the Cornell News site.
The models that have been developed to Times Square, the shopping district Akhibahara in Tokyo and 5Pointz a graffiti exhibition area in Queens, then a viewer can navigate within a virtual 3-D space while using a slider to move forward and backward in time and viewing reality in various positions.

The software works with flat surfaces in the image: signs, walls, façades or panels theater and then treats as “patches” which are stitched together to create the overall scene, the 3D effect, so it is made from a flat stage and need not be shot in 3D original, facilitating the shooting and the size of the files.

The 3-D model is then created from the global scene at a particular moment in time, the computer joins the slices that are taken from each solid in the right time and reconstructs the image for each time more simple and functional way.

The researchers who presented the model at the fair in Zurich Computer Vision, said: “when we reflect on the time of image collections, each observation has something to say,” they received the award for Best Book in the area last year.

They figure that may also change the way we think about photography, they said, and some day all photo you thought about doing, may have been made by someone.

 

(Português) Equipe recorre a torcida de robôs

07 Aug

Sorry, this entry is only available in Brazilian Portuguese.

 

(Português) As belas ondas de WiFi

22 Jul

Sorry, this entry is only available in Brazilian Portuguese.

 

(Português) Termina em São Paulo feira de segurança

16 May

Sorry, this entry is only available in Brazilian Portuguese.