Google Machine Learning APIs

As seen in the previous chapter, machine learning is used in a wide variety of applications. However, a few applications are easy to build, while a few are very hard to build, especially for a user who is less familiar with machine learning. Some of the applications that we are going to discuss in this chapter fall in the hard to build category, as the process of building a machine learning model for these applications is data intensive, resource intensive, and requires a lot of knowledge in the field.

In this chapter, we will go over five machine learning APIs provided by Google (as of March 2018). These APIs are meant to be used out of the box, as RESTful APIs. For each service mentioned in the following, we will show what type of application can benefit from it, and how to interpret the returned results:

• Vision has a label detection, OCR, face...

Vision API

The Vision API lets us build quite a few applications related to vision:

• Detecting labels in an image
• Detecting the text in an image
• Face detection
• Emotion detection
• Logo detection
• Landmark detection

Before we dive into building applications using the preceding, let's get a quick understanding of how they might be built, using face emotion detection as an example.

The process of detecting emotions involves:

1. Collecting a huge set of images
2. Hand-labeling images with the emotion that is likely represented in the image
3. Training a convolutional neural network (CNN) (to be discussed in future chapters) to classify the emotion, based on an image as input

While the preceding steps are heavily resource intensive (as we would need a lot of humans to collect and hand-label images), there are multiple other ways to obtain face emotion detection. We are not sure how Google...

Cloud Translation API

The Cloud Translation API provides a simple, programmatic interface for translating an arbitrary string into any supported language, using state-of-the-art neural machine translation. The Translation API is highly responsive, so websites and applications can integrate with the Translation API for fast, dynamic translation of source text from the source language to a target language (for example, French to English). Language detection is also available for cases in which the source language is unknown.

Enabling the API

For us to be able to use Google cloud translation services, we need to enable, which is done as follows:

1. In order to enable the Google Cloud Translation API, search for the API in the console...

Natural Language API

The Google Cloud Natural Language API reveals the structure and meaning of text by offering powerful machine learning models in an easy-to-use REST API. You can use it to extract information about people, places, events, and much more, that are mentioned in text documents, news articles, or blog posts. You can also use it to understand the sentiment about your product on social media, or to parse intent from customer conversations happening in a call center or a messaging app. You can analyze the text uploaded in your request, or integrate it with with your document storage on Google Cloud storage.

The Cloud Natural Language API can be found by searching for it in your console, as follows:

The Cloud Natural Language API is enabled in the resulting page:

Similar to the Translation API, we do not have to create credentials for this API if at least one API...

Speech-to-text API

The Google Cloud Speech API enables developers to convert audio to text, by applying powerful neural network models in an easy-to-use API. The API recognizes over 110 languages and variants. One can transcribe the text of users dictating to an application's microphone, enable command-and-control through voice, or transcribe audio files, among many use cases.

In order to enable the speech to text API, search for it in the console, as follows:

In the resulting web page, enable the API, as follows:

Similar to the APIs mentioned in the previous sections, credentials obtained for one API can be replicated for the other Google APIs. So, we don't have to create credentials separately for the speech to text API.

Once the API is enabled, let's start the Cloud Shell and Datalab, as we did in the previous sections.

In the following code, we transcribe...

Video Intelligence API

The Cloud Video Intelligence API makes videos searchable and discoverable, by extracting metadata with an easy-to-use REST API. You can now search every moment of every video file in your catalog. It quickly annotates videos stored in Google Cloud storage, and helps you to identify key entities (nouns) within your videos and when they occur.

The Cloud Video Intelligence API can be searched for and enabled as follows:

We import the required packages and add the path to the api-key, as follows:

from google.cloud import videointelligence
import os
os.environ["GOOGLE_APPLICATION_CREDENTIALS"] = "/content/datalab/google-api.json"
from google.cloud.speech import enums
from google.cloud.speech import types

The method features enables us to specify the type of content that we want to detect in a video. The features available are as follows:

Let...

Summary

In this chapter, we went through the major machine learning APIs that Google provides: vision, translate, NLP, speech, and video intelligence. We have learned how the various methods in each of the APIs enable us to replicate deep learning results, without having to code from scratch.