What is Deep Learning?
Deep learning is a type of machine learning that uses artificial neural networks to enable digital systems to learn and make decisions based on unstructured, unlabeled data.
Generally, machine learning trains AI systems to learn by examining experiences with data, recognizing patterns, making recommendations, and adapting. Especially when it comes to deep learning, digital systems learn from examples rather than just responding to rule sets, and then use that information to react, behave, and perform like humans.
Tensorflow is an open-source library for numerical computation and large-scale machine learning that ease Google Brain TensorFlow, the process of acquiring data, training models, serving predictions, and refining future results.
Tensorflow bundles together Machine Learning and Deep Learning models and algorithms. It uses Python as a convenient front-end and runs it efficiently in optimized C++.
Tensorflow allows developers to create a graph of computations to perform. Each node in the graph represents a mathematical operation and each connection represents data. Hence, instead of dealing with low-details like figuring out proper ways to hitch the output of one function to the input of another, the developer can focus on the overall logic of the application.
The deep learning artificial intelligence research team at Google, Google Brain, in the year 2015 developed TensorFlow for Google’s internal use. This Open-Source Software library is used by the research team to perform several important tasks.
TensorFlow is at present the most popular software library. There are several real-world applications of deep learning that makes TensorFlow popular. Being an Open-Source library for deep learning and machine learning, TensorFlow finds a role to play in text-based applications, image recognition, voice search, and many more. DeepFace, Facebook’s image recognition system, uses TensorFlow for image recognition. It is used by Apple’s Siri for voice recognition. Every Google app that you use has made good use of TensorFlow to make your experience better.
TensorFlow's open source models
The TensorFlow team has open sourced a large number of models. You can find them in the tensorflow/models repo. For many of these, the released code includes not only the model graph, but also trained model weights. This means that you can try such models out of the box, and you can tune many of them further using a process called transfer learning.
Here are just a few of the recently released models (there are many more):
The Object Detection API: It's still a core machine learning challenge to create accurate machine learning models capable of localizing and identifying multiple objects in a single image. The recently open sourced TensorFlow Object Detection API has produced state-of-the-art results (and placed first in the COCO detection challenge).
Main Use Cases of TensorFlow
One of the most well-known uses of TensorFlow are Sound based applications. With the proper data feed, neural networks are capable of understanding audio signals. These can be:
Voice recognition – mostly used in IoT, Automotive, Security and UX/UI
Voice search – mostly used in Telecoms, Handset Manufacturers
Sentiment Analysis – mostly used in CRM
Flaw Detection (engine noise) – mostly used in Automotive and Aviation
Regarding common use cases, we are all familiar with voice-search and voice-activated assistants with the new wide spreading smartphones such as Apple’s Siri, Google Now for Android and Microsoft Cortana for Windows Phone.
Language understanding is another common use case for Voice Recognation. Speech-to-text applications can be used to determine snippets of sound in greater audio files, and transcribe the spoken word as text.
Sound based applications also can be used in CRM. A use case scenario might be: TensorFlow algorithms standing in for customer service agents, and route customers to the relevant information they need, and faster than the agents.
2-Text Based Applications
Further popular uses of TensorFlow are, text based applications such as sentimental analysis (CRM, Social Media), Threat Detection (Social Media, Government) and Fraud Detection (Insurance, Finance)
Language Detection is one of the most popular uses of text based applications.
We all know Google Translate, which supports over 100 languages translating from one to another. The evolved versions can be used for many cases like translating jargon legalese in contracts into plain language.
Google also found out that for shorter texts, summarization can be learned with a technique called sequence-to-sequence learning. This can be used to produce headlines for news articles. Below, you can see an example where the model reads the article text and writes a suitable headline.
Another Google use case is SmartReply . It automatically generates e-mail responses (wishing for the evolved version of this one doing our business on behalf of us)
Mostly used by Social Media, Telecom and Handset Manufacturers; Face Recognition, Image Search, Motion Detection, Machine Vision and Photo Clustering can be used also in Automotive, Aviation and Healthcare Industries. Image Recognition aims to recognize and identify people and objects in images as well as understanding the content and context.
TensorFlow object recognition algorithms classify and identify arbitrary objects within larger images. This is usually used in engineering applications to identify shapes for modeling purposes (3D space construction from 2D images) and by social networks for photo tagging (Facebook’s Deep Face). By analyzing thousands of photos of trees for example, the technology can learn to identify a tree it has never seen before.
Image Recognition is starting to expand in the Healthcare Industry, too where TensorFlow algorithms can process more information and spot more patterns than their human counterparts. Computers are now able to review scans and spot more illnesses than humans.
TensorFlow Time Series algorithms are used for analyzing time series data in order to extract meaningful statistics. They allow forecasting non-specific time periods in addition to generate alternative versions of the time series.
The most common use case for Time Series is Recommendation. You’ve probably heard of this use from Amazon, Google, Facebook and Netflix where they analyze customer activity and compare it to the millions of other users to determine what the customer might like to purchase or watch. These recommendations are getting even smarter, for example, they offer you certain things as gifts (not for yourself) or TV shows that your family members might like.
The other uses of TensorFlow Time Series algorithms are mainly the field of interest to Finance, Accounting, Government, Security and IoT with Risk Detections, Predictive Analysis and Enterprise/Resource Planning.
TensorFlow neural networks also work on video data. This is mainly used in Motion Detection, Real-Time Thread Detection in Gaming, Security, Airports and UX/UI fields. Recently, Universities are working on Large scale Video Classification datasets like YouTube-8M aiming to accelerate research on large-scale video understanding, representation learning, noisy data modeling, transfer learning, and domain adaptation approaches for video.
Bonus: This might not be a common use but it is a matter of life and death –if you watch American movies you know! – and it is chosen as the top 20 projects worldwide (Global Finalist). Nasa is designing a system with TensorFlow for orbit classification and object clustering of asteroids. As a result, they can classify and predict NEOs (near earth objects).
As TensorFlow is an open source library, we will see many more innovative use cases soon, which will influence one another and contribute to Machine Learning technology.
Other companies listed on the TensorFlow website as users of the framework include Airbnb, Coca-Cola, eBay, Intel, Qualcomm, SAP, Twitter, Uber and Snapchat developer Snap Inc. Another user is STATS LLC, a sports consulting company that runs TensorFlow-based deep learning models to analyze things such as the movements of players during professional sports games.
TensorFlow-based deep learning has also been a part of experiments and tests involving one of the larger-scaled proposed innovations today, that is self-driving cars.
Some smaller-scale uses have been found, too. For example, a small Japanese farm uses TensorFlow to sort cucumbers based on their textures.
Some Other Examples: