Google Pixel Event: Transition from “mobile first” to “AI first” strategy

On 4th October, Google announced a range of hardware at a special event conducted in San Francisco. Artificial intelligence is the most important development that takes place in the consumer technology space. Google’s equation for new products seems to reflect the same – software + hardware and then integrate it with AI. Here are some important highlights from the event:

Google Announced Pixel 2 smartphone – a 5 inch and a 6 inch display version. The phone is powered by a 12 Megapixel rear cam and an 8 megapixel front cam. There are a lot of exciting features in Pixel 2 like activating Google Assistant by squeezing your phone, Google Photos capabilities including Lens, as well as virtual- and augmented-reality capabilities and finally you have a male voice to Google Assistant. The Pixel 2 phone will be available in three colors – “Just Black,” “Clearly White” and “Kinda Blue”. The 5-inch Pixel 2 starts at $649, while the larger 6-inch Pixel 2 XL will cost at least $849.

The most interesting announcement that we felt interesting was the Google Pixel Buds – a wireless headphones that is going to rival Apple’s Bluetooth-powered AirPods. The most interesting feature is that it can do an instant translation for a variety of languages. When the user speaks her native language, an English translation is provided from the phone’s speakers. So, next time if you stuck in another country, just put up your Pixel Buds and start conversing. The Pixel Buds come in the same three colors as the Pixel phone and cost $159.

Google also debuted the Pixelbook – a brand new laptop powered by Chromebook. It has a 12.3 inch screen that’s meant to function both as a laptop and a tablet. Priced at $999, it’s also the first laptop to have Google Assistant. You can also purchase a stylus for $99 that lets you interact with your laptop.

Google also announced Home speakers – Google Home Mini, a smaller version of Google Home speaker and a high-end smart speaker called Google Home Max. Google claims that the Home Max offers deeper bass powered by dual 4.5-inch woofers, and it’s 20 times more powerful than the original Google Home. It shows that Google clearly intends to compete with Apple’s HomePod and Amazon’s new Echo. The Google Home Mini will cost $49 whereas Google Home Max will cost $399.

They also showed off called Google Clips – a new clip-on camera that can take photo and video remotely and synchronizes what it captures instantly with a connected phone through an app. Well, GoPro’s share did fall when the announcement was made.

Google also redesigned the virtual-reality headset Daydream view that works with the Pixel smartphone. It has better lenses that will increase the image clarity and field of vision.

If we look at brand new Pixel 2 phones, Google Home Mini, Google Home Max, Google Pixel Buds, and Google Clips camera, everything is powered by Google Assistant. It’s clear, the future is going to be AI driven and Google has already made its moves.

Another interesting point to observe is that there were more female presenters for the event. There were four when compared to last year’s two. Google was embroiled in controversy earlier this year when former company engineer James Damore wrote a manifesto arguing that women are underrepresented in tech because of “biological causes” rather than bias and discrimination. Looks like Google is just getting started.

Getting to know Deep Learning with TensorFlow

Machine learning is one of the fastest-growing and most exciting fields in the 21st Century. It is the science of getting computers to act without being explicitly programmed. Deep Learning is a subfield of machine learning that is concerned with algorithms that is inspired by the structure and function of the brain called artificial neural networks.

Today, most companies hire talents in the field and develop Machine learning tools & packages to develop products/services. Google is one of the major companies that uses machine learning extensively. Google uses machine learning to classify images in Google Photos, for image recognition, voice recognition, to improve search results, translation and search suggestions.

For developers, Google has developed an open-source software machine learning library called TensorFlow. It was originally developed by the Google Brain team for their internal use. Google Brain developed DistBelief – a proprietary machine learning system based on deep learning neural networks. The use of Distbelief grew rapidly with the expansion of the company. TensorFlow was Google Brain’s second generation machine learning system and it was released as open source software on November 9, 2015. TensorFlow Lite is a scaled-down version of the software library meant to run on mobile devices.

What is a Tensor ?

The central unit of data in TensorFlow is the tensor. The name TensorFlow derives from the operations in which neural networks that performs on multidimensional data arrays. A tensor consists of a set of primitive values shaped into an array of any number of dimensions. A tensor’s rank is its number of dimensions. So, basically it’s a multi-dimensional array.

So, why don’t they call it a multi-dimensional array ? Here is where you must understand the difference. Tensors and multidimensional arrays are different types of object; the first is a type of function, the second is a data structure suitable for representing a tensor in a coordinate system.

In short, Tensor is “something” which can be represented as multidimensional array. This representation must depend in some very specific way from where you are looking at this “something”.

To get started with TensorFlow, head towards the TensorFlow website that has a bit of documentation and tutorials. Please note that these are made on the assumption that you have some (or a lot) prior machine learning and artificial intelligence knowledge/experience. You also should be very comfortable with Python programming language.

Udacity has partnered with the Google to provide a course on Deep Learning. The course is available for free and TensorFlow will be introduced via projects and assignments. For more details visit: Deep Learning with Google

There are lot of exciting project that have been developed using TensorFlow. Visit the collection of TensorFlow models in the TensorFlow Zoo.

One of the main reasons why Google open-sourced TensorFlow is to attract the best talent from the world. Many top professors from universities have joined companies like Facebook, Google & Baidu. Google has certainly a first mover advantage in terms of open-source TensorFlow. Apple also generate huge amount of data, but their extreme stance on privacy has restricted the development. Google is acquiring a lot of startup/companies based on Machine Learning and AI, the recent Halli Labs from India, data science community Kaggle & DeepMind in 2014. Looks like Google is now becoming an A.I. first company.

TensorFlow is already one of the leading platforms for teaching A.I. and the adoption rate is expect to grow exponential. So, if you are looking for a machine learning job, start working on project with TensorFlow.

Android O: All you need to know about it

Google unveiled Android O at the Google I/O annual developer conference held on 17-19th May 2017. Google has released the second developer preview of Android O. In case you missed, we have the important highlights for you about the upcoming update:

Battery life has been one of the major concerns for any android customer. Google is trying it level best to limit background activity and maximize battery life. In the latest upcoming update, it has increased the automatic limits on what apps can do in the background in a number of key areas (broadcasts, background services, location updates).

Google is bringing new PiP (picture-in-picture) features that enable users to continue watching videos while working within other apps; apps will be able to put themselves in PiP mode. Something similar you can find in Facebook app, where you can continue to watch a video, looking at other posts.

Google is also introducing new Notification channels that allows to group notifications together by their type. Users can block or change the behavior of each channel individually, rather than managing all of the apps’ notifications together. It will help you things like how a news app notifies us or a music player shows a persistent notification.

Another new feature is notification dots, which are visual indicators on app icons that’ll show if you have any waiting notifications. iOS users maybe familiar with this.

AutoFill API’s will help you remember your pre-filled form, usernames/passwords (in some cases) to quickly be added into apps on your device. A user will be able to choose a source for autofill data, and applications that need to store and retrieve this sort of data no longer will need to act as an Accessibility service.

Android on Chromebooks also indicates that there will be an update to the keyword navigation. Android O focuses on building a better model for arrow and tab key navigation.

With the new updates, Google has also announced Adaptive icons and badges. The feature will enable developers to use different-shaped app icons depending on the manufacturer’s preference. Adaptive icons will be supported in the launcher, shortcuts, device Settings, sharing dialogs, and the app overview screen.

Wi-Fi Awareness will allow your Android O device or app to communicate with other devices and apps in the vicinity over Wi-Fi without requiring an actual internet connection. Sony’s LDAC codec is also part of the update. They are also introducing the AAudio, which could result in improved low-latency audio.

From the latest update, the Pandas feel that the Google is following Nougat’s footstep. The first developer preview will be available on the Nexus 5X, Nexus 6P, Nexus Player, Pixel, Pixel XL and Pixel C devices. Google is concentration more on improving performance rather than just adding features. It’s focusing on strengthening the platform and filling gaps. We are eagerly waiting for the update and probably hear lots more about Android. Google is also yet to name the update – Have any suggestions ? – We would like to hear from you from the comments below

Stay tuned for the updates

Amazon Fire vs Google Chromecast

On 19th April 2017, Amazon launched its Fire TV Stick in India. Both Amazon Fire Video Stick and Google Chromecast are both plug and play devices priced at Rs 3,999 (Rs 499 cashback via Amazon Pay for existing Amazon Prime customers). On Between the Google Chromecast 2 and Amazon Fire TV Stick, which one is a better buy? Let’s find out which one is better !

Amazon Fire Stick is a plug and play device that turns any HD television set in to a smart TV that streams movies, lets you use apps and play games, among others. You can plug in the device into a TV with a USB support and view videos from Amazon Prime Video, NetFlix, YouTube etc. One of the most interesting features of Fire Stick is the remote which allows voice control.

The company has also tied up with Airtel for providing 100 GB of data free to Airtel home users purchasing a Fire TV Stick. There’s also a three-month free premium subscription for Eros Now, six months of ad-free streaming on Gaana, and a month of ad-free streaming on Voot.

Fire TV Stick which streams content through WiFi, whereas Google Chromecast 2 mirrors videos, music, and apps from your smartphone. It is also a plug and play device and mirrors contents from your smartphone (you need one) to your TV. You can download apps via Play Store and view contents on your TV. Therefore, the functionality is limited – you may not be able to use your phone to make or receive calls. Streaming content for long hours also drains your phone’s battery.

In terms of technology innovation, Fire TV Stick is capable of ‘ASAP’ i.e. advanced streaming and prediction, where the device learns what you a user watches and starts pre-fetching to reduce wait times. The company claims that this can speed up video playback by 50 percent. It also has support for the more demanding H.265 (HEVC) video codec, which means movies will use less data and take less time to buffer.

The Pandas feel that Fire TV stick is great if you’re a Prime Video subscriber who also uses Netflix and YouTube. But looking at the Indian scenario, there needs to be more support for Indian apps. But any product in its initial days would have some limitations, and over time with updates, it is expected to fix bugs and bring more improvements.