FacebookGoogle PlusLinkedIn

Archive

Posts Tagged ‘developer’

Amazon and Microsoft Combine Powers to form – GLUON! AI Deeper Learning on Neural Networks!

Monday, April 23rd, 2018

In the highly competitive world that we live in today, its clear to understand that there is a strict competitive market building amongst the great tech companies such as Microsoft, Google, Amazon, Baidu, and other startup companies. All these companies have developed, deployed, and integrated their cloud based solutions with Artificial Intelligence. Artificial intelligence provides the steroids that the Cloud industry needs in order to accelerate its existence, growth, and empowerment. AI has developed key components to help itself with its constant learning processes. These processes can be better understood through “machine learning”. Machine learning allows information to be rapidly digested, conceptualized, and intelligently categorized by artificial intelligence. Cloud products such as Azure from Microsoft, AWS from Amazon, TensorFlow from Google, and more integrate machine learning into their systems to enhance the user experience and make mundane tasks to now be automated with little or no human effort involved.

With this evolution and the stiff competition there can be no collaboration in sight especially from rival companies which have no need to partner up, but to simply dominate their place in the industry. Surprisingly Microsoft and Amazon have done such a collaboration. This collaborated effort is not only to show that larger companies can work together, but also the data collected from this program allows both companies to be more successful in their domination of the cloud space.  Microsoft and Amazon have come together to form Gluon. Gluon in the dictionary is defined as a “subatomic particle of a class that is through to bind quarks together”. However, this is not the same Gluon which are referring to. Gluon is the fist distinctive collaborative efforts of these technology giants in the race for Artificial Intelligence and gathering information. Gluon is considered the store-house for machine learning and the ability to “voluntarily” utilize machine learning in the development of products and services by Amazon and Microsoft. This storehouse of technological learning and a knowledge base that is bubbling over with information is the type of data warehouse that AI thrives on.

This cohesive collaboration between developers and machine learning have no blossomed what is dubbed “deep learning”. Deep learning essentially is a combination of three distinct components such as data for training, neural network model, and an algorithm which trains the neural network.  The neural network essentially translates the data and feeds AI this information allowing AI to grow on a more diversified scale.  This new algorithm that is now being utilized by neural networks self-adjusts its output based on errors in the network output. This is a memory and compute process that machine learning and AI adapts through predictive outputs. An example of deep learning is Caffe2, TensorFlow, Appache MXNet, and Cognitive Toolkit offers options to speed up the neural network which often takes days to compute the data being derived.  These products now reduce the learning time and accelerates the parallelization in distributing computation processes.

Even though these products built within Gluon are effective and pose to change the way AI is developing by giving it more knowledge, the key is having developers to utilize these products to diversify and intensify the data stream.  AWS has been experimenting with its developers by using MXNet to train the neural networks. Microsoft has become a heavy contributor the open source MXNet by opening it to an ever-increasing rise in developers. The collaboration and data being created and distributed by these two powerhouses with Gluon can be overwhelming for a beginner when first interacting with this program, but even for more advanced developers, the data intensive algorithms seem take a life of its own demanding more ways to conform and adjust with massive error reduction.  The four key innovations which is introduced by Gluon are as follows:

Friendly API – using clear and concise code, it allows developers to learn and understand the data.

 

Dynamic networks – allows for ease of access and the rapid fluctuation of the data structure. Fluency in the data structure is critical for development as it allows hybrid scenarios and reduces stagnancy of the data flow as with previous machine learning software.

 

Algorithms that devein the network – Seamless combination of the model and algorithm allows the network to adjust definitions during training. This is critical as it allows developers to use programming loops, and conditionals.  Algorithms are now easier to change, create, and debug.

High Performance Operators for training – Gives the ability to create dynamic graphs and concise APi without sacrificing speed. Previous versions seemed to have consumed valuable run time that this feature runs through effortlessly and drastically picks up speed.

The question now becomes, how does developers access Gluon? Well its provided through Apache MXNet and the future releases will have support for Microsoft Cognitive Toolkit.  The AWS team has already published a front-end interface with low-level API to include other specifications and frameworks. Once accessing Gluon you can utilize an AWS Deep learning AMI to find a plethora of examples and workbooks utilized and documented by fellow developers.

 

Could this be a taste of what is to come from these super giant tech companies in terms of collaboration in order to leap ahead or is this just an opportunity for the public to advance an already thriving technology called Artificial Intelligence? These tech giants are clearly leveraging all aspects of “free data” and utilizing the voluntary efforts of developers world-wide to feed this neural network. As a partner for Microsoft, we anticipate that Microsoft as being the leader in products on the market will utilize Gluon in order to create more intensive, responsive, and advanced products built with AI as the backbone to reduce many processes and essentially speed up productivity. It Gurus Of Atlanta is your Microsoft partner of choice which brings you the cutting edge in design, technology, and its advancements.

“All Your Company Information Technology Needs Under One Company”

3355 Lenox Road Atlanta, GA 30326

www.itgurusatl.com| customerservice@itgurusatl.com|

(888) 511-0143 OR (706) 406-5914

“As a registered SAM.gov company, we service government entities across the entire US and Canada”

“LIKE” us on Facebook at: www.facebook.com/itgurusatl / “FOLLOW US” on Twitter: www.twitter.com/itgurusatl / “CONNECT” with us on LinkedIn: www.linkedin.com/in/itgurusofatlanta/

SUBSCRIBE TO OUR NEWSFEED AND UPDATES AT: http://eepurl.com/cU_t7r

 

Microsoft Azure, Machine Learning, and Artificial Intelligence! – LUIS

Wednesday, February 28th, 2018

Azure, which is Microsoft’s great platform for virtualizing an Active Directory environment, or at least that is what it started out being. Since its introduction in October of 2008 and its release into the production for masses on February 1st, 2010, Azure has grown into a lot more than just a virtual Active Directory. Azure has now become the world’s largest platform for virtual technology. This includes app development, deployments, and the introduction of PaaS, IaaS, and SaaS. These technological advancements within the Azure space allows Azure to not only be manipulated, but also enhanced to include Machine Learning.

Machine Learning is the next phase in technological revolution. Machine Learning gives the life to systems allowing them to self-correct and self-rewrite their programming to be more fundamentally correct. Machine Learning is already in most of the products that we use daily such as Amazon’s Echo which utilizes speech to determine products and services that best suite you, the consumer.  Other examples are Siri within the iPhone, Cortana with Windows, and Google Assistant. All these technologies work together with a simple purpose. That purpose is to learn the habits and functions of human beings to become better and more sufficient.

The connection flag ship with allows all these devices and machine learning products to interface is called IoT or the Internet Of Things. IoT connects devices such as your smart phone, home security, bluetooth enabled appliances, and more on to one unified management platform that they call interconnect and they are easier to manage. IoT is the next wave in technological advancement being presented by Microsoft along with other large competitors such as Amazon and Google. IoT is becoming the next big thing when it comes to full home automation. Machine learning interprets all information being gathered from analytics, bots, IoT experiences, including conversational components into a spoken dialogue called LU. LU or Language Understanding determines the intent of a sentence including the machine-enabled meaning representation.

In an effort in increase Machine Learning, Microsoft has launched LUIS which is its Language Understanding Intelligent Service. LUIS will allow Software developers to create cloud-based machine learning LU models that are designed around their specific application domains. This allows the code to be written and interpreted with very little Machine Learning experience. LUIS is now a cloud-based application which does most of the background work for Machine Learning in the background while giving users the agility and flexibility to cater the experience more towards their application and organization.  The application uses user experiences from the developer and learns how to obtain an HTTP endpoint in Azure which will then receive real-time traffic.  This is called “Active Learning”.  Through these “Active Learning” utterances, LUIS identifies the key features needed to make the experience and solution unique to that user.  These active learning sessions are done until optimum levels of requirements are met. Bear in mind that active learning is also utilizing machine learning, so the learning curve is very small, while the results are extraordinarily accurate.

LUIS runs off of three main functions and they can be categorized as Intents, Utterances, and Entities. All three aspects engage LUIS functionality and ability to learn the user habits. A more detailed overview of these functions are listed below:

 

Intents: The input done by a user to express actions with a purpose that the user wishes to perform or a goal which the user is trying to accomplish.  This could be as simple as booking flight, hotel, or pulling up a newspaper article. The name of the action can then be associated with the action.

 

Utterances: This can be a combination of text inputs that regulate an action to obtain results. Such actions in a text can be “check flight status” or “what was the score from the game?”.  Because of so many variations to an utterance, they may not always be perfectly formed sentences, but all desire a intent.

Entities: An entity is a specific detail within an utterance which governs the direction the utterance will take.  An example of this is “hotels in Jamaica”. “Jamaica” is the location. LUIS can determine the location, then using the other intents to understand the utterance, then provide a response.

 

LUIS utilizes powerful entity extractors for it to achieve learning capabilities and become more successful with its responses. LUIS allows developers to quickly build language understanding applications. The applications can then combine with customizable pre-built apps to include music, dictionaries, calendars, and devices. Through interaction with the developer and the information being constantly pulled from the internet, the learning, and solutions that LUIS provides becomes more intuitive with each use.  Once the applications are created using LUIS, the app can then be customized and tailored to the users which the app is designed for giving all of them that unique experience.

LUIS uses two ways to build a model, that is through “LUIS.ai web app” and “Authoring APIs”. Whether the user goes with the LUIS.ai web app or the Authoring APIs, they both give control over the LUIS model definition.  Another fundamental and creative method that developers have found is to combine both to build a model.  Management within the model includes models, versions, external APIs, collaborations, training, and testing.

LUIS is another segment in the Machine Learning process in the revolution of Artificial Intelligence. As machines begin to learn increasingly about human behavior and what we like, they also become aware of errors that we make including how to fix or to alleviate the errors all together.  Error free, automation, increase in productivity, and cutting cost is the motive behind AI. The Microsoft Azure cloud space is proving to be a formidable place for this technology to not only thrive but to maximize on limitless possibilities. IT GURUS OF ATLANTA will ensure that we provide updates to our supporters as we are a trusted Microsoft Partner and a certified Microsoft Cloud Partner.

“All Your Company Information Technology Needs Under One Company”

3355 Lenox Road Atlanta, GA 30326

www.itgurusatl.comcustomerservice@itgurusatl.com|

(888) 511-0143 OR (706) 406-5914

“As a registered SAM.gov company, we service government entities across the entire US and Canada”

“LIKE” us on Facebook at: www.facebook.com/itgurusatl / “FOLLOW US” on Twitter: www.twitter.com/itgurusatl / “CONNECT” with us on LinkedIn: www.linkedin.com/in/itgurusofatlanta/

SUBSCRIBE TO OUR NEWSFEED AND UPDATES AT: http://eepurl.com/cU_t7r

 

Citizenship to a Robot? Yes, Sophia is the World’s First Robot With Citizenship Rights!

Friday, February 9th, 2018

By now many reading this article are thinking in their heads, what is Sophia and how come I have never heard of her? Most importantly what and why would a robot attain citizenship. Well IT GURUS OF ATLANTA is here to tell the world that the citizenship is no joke. As of October 25th, 2017, Sophia which is a full-sized robot which has demonstrated social skills and knowledge unmatched by many others in the world.  Sophia of course by the name is a female robot and is also a full pledged citizen of Saudi Arabia which granted the citizenship to the robot.

Sophia is a perfect example of the work of Artificial Intelligence or best called AI. Sophia is capable of speaking concerning business across multiple platforms and industries. She has already met with a variety of decisionmakers across the world and assisted with many different key business decisions. The United Nations has already recognized Sophia by giving her an official role working with UNDP to safeguard human rights. How can a robot accomplish so much in such a little time, well its precisely said, Sophia is a robot and is currently the flagship for her maker, which is Hanson Robotics.

Sophia has managed to grab a reader base that subscribe to her videos and press coverages of over 10 billion readers in 2017. This is an astronomical figure as most celebrities themselves do not reach a pinnacle of that much of a fan-based audience.   The key feature of Sophia is her ability to display emotion and adapt to her environment by changing the tone, then matching what she is saying with the appropriate expression. The creator and owner of Sophia is Dr. David Hanson. Mr. Hanson worked previously as an “Imagineer” at Disney.   Because of his genius he was able to leave Disney and create his own company. Mr. Hanson’s motto is that to for robot to have the fundamental likeness of humanity, then the robot should possess creativity, empathy, and compassion. Those three traits are the design for a fully interactive experience with an AI such as Sophia.

 

The next phase in the Hanson Robotics project is to manufacture Sophia or “like-Sophia’s” for home use. Currently the production cost of the original Sophia is too much for the average person to afford, but the open platform allows for any developer to create their own Sophia built off of the current algorithms that Sophia is running on. Hanson Robotics which is based out of Hong Kong, built Sophia and modelled her features off of Audrey Hepburn, whom was a famous British actress.  Much of the features of crying. Laughing, and even expressing joy or anger is features that were displayed by Audrey Hepburn during her time.

Sophia has graduated since her original introduction to now being able to walk on her own with Hanson Robotics adding legs to her repute.  As AI technology develops so does the learning capabilities and functions with Sophia. She is adamant about protecting human rights and is a clear derivative of the future and where we as a society is heading. Now how much control and human-likeness will be give this humanoid generation is unclear, but definitely is a eye-brow raiser as the future begins to unfold. IT GURUS OF ATLANTA is dedicated to ensure that these technological breakthroughs are brought to the cusp of inquiring minds. It is our dedication and quality that recognizes these changes and adapt along with the change of times.

 

“All Your Company Information Technology Needs Under One Company”

3355 Lenox Road Atlanta, GA 30326

www.itgurusatl.comcustomerservice@itgurusatl.com|

(888) 511-0143 OR (706) 406-5914

“As a registered SAM.gov company, we service government entities across the entire US and Canada”

“LIKE” us on Facebook at: www.facebook.com/itgurusatl / “FOLLOW US” on Twitter: www.twitter.com/itgurusatl / “CONNECT” with us on LinkedIn: www.linkedin.com/in/itgurusofatlanta/

SUBSCRIBE TO OUR NEWSFEED AND UPDATES AT: http://eepurl.com/cU_t7r

 

China now using Face-Recognition software to grant access to facilities and Track criminals

Saturday, July 22nd, 2017

 

Companies like Face++ are changing the landscape in China when it comes to technology and utilizing facial-recognition software. The company is a startup company, but is already valued at over a billion dollars. Face++ is using its investments in order to diversify and integrate facial-recognition software into critical security systems.

The software updates done by Face++ are cutting edge which allows devices to track 83 facial points at the same time, and from multiple angles to give recognition ability which is like no other.  The updates that Face++ has been giving facial-recognition is already being integrated in many popular applications such as Alipay and Didi. Both applications are major applications in China that are used to make mobile payments and get rides like Uber has done in the US. The Didi app allows users to double check the driver that is coming by making them move their heads and use speech recognition that no one can substitute the driver or use a photo. Payments via the Alipay app allows users to use face and voice combinations in order to complete payments.

 

The Chinese government has already invested using Face++ facial-recognition software to tap into surveillance being done on criminals by tapping into their centralized ID databases and cameras to quickly apprehend and document activity by known criminals. Even though the mugshots and the images obtained from cameras are not all the time accurate, it gives a percentage possibility of that person being the person of interest and narrows down the possibility of monitoring or apprehending the wrong suspect. Utilizing cameras and face-recognition is definitely ahead of its time and is geared to change not only the Chinese government’s methods of fighting crime, but will soon spread to other countries as it gains more popularity in the security front.

 

 

 

Behind the nuts and bolts of Face++ facial-recognition software is their innovative artificial intelligence which is constantly getting better with each use. Deep learning is dubbed the brain-power behind the facial-recognition software that depicts the motions and movements of the face from infinite angles and positions to categorize the facial expression possibilities. Deep learning bridges the gap between AI and machine learning. With Deep Learning the capabilities are endless by utilizing what is dubbed “neural networks” which essentially translates human related data into machine learning for AI to utilize.

 

IT GURUS OF ATLANTA LLC

“All Your Company Information Technology Needs Under One Company”

3355 Lenox Road Atlanta, GA 30326

www.itgurusatl.comcustomerservice@itgurusatl.com|

(888) 511-0143 OR (706) 406-5914

“As a registered SAM.gov company, we service government entities across the entire US and Canada”

“LIKE” us on Facebook at: http://www.facebook.com/itgurusatl / “FOLLOW US” on Twitter: http://www.twitter.com/itgurusatl / “CONNECT” with us on LinkedIn: http://www.linkedin.com/in/itgurusofatlanta/

SUBSCRIBE TO OUR NEWSFEED AND UPDATES AT: http://eepurl.com/cU_t7r


Call Now To Get Started!

(888) 511-0143 or (706) 406-5914

Contact Us
close slider

Contact Form

Fields marked with an * are required