Machine Learning in Google Cloud Platform

From Luis Gallego Hurtado - Not Another IT guy
Jump to: navigation, search

Google offers different Artificial Intelligence / Machine Learning products as a service.

Google Vertex AI

Google Vertex AI is the unified platform to help you build, deploy and scale more AI models

  • Prepare and store your datasets
  • Experiment and deploy more models faster.
  • Manage your models.
  • Pre-trained APIs for vision, video, natural language and more.
  • Integrated with other Google Services. Through Vertex AI Workbench, Vertex AI is integrated with DataFlow for feature processing, Cloud Storage for storage, and DataLab for model creation.
  • Portable models with open source Tensor Flow, PyTorch and scikit-learn.

Google Cloud Machine Learning allows you to build scalable machine learning models, based in a Tensor Flow implementation.

  • Scalable and managed service.
  • You can discover and share samples.
  • Hypertune (hyperparemeter tuning).
  • You can create and analyse models with Notebook Developer Experience.

AutoML Video

AutoML Video makes videos searchable and discovereable, extracting metadata.

  • Getting insight from videos
  • Label detection, annotating videos, identify key nouns entities in your videos.
  • Change shot detection, recognizing when labels occur.
  • Integrated with other Google Cloud services, like Cloud Storage.

AutoML Image

AutoML Image helps you to understand the content of an image.

  • Label detection, classifying images.
  • Explicit content detection.
  • Logo detection.
  • Landmark detection.
  • Optical character detection: it finds and reads words from images.
  • Face detection.
  • Web detection: similar images in web.
  • Image attributes.
  • You can build metadata, moderate offensive content, or run sentiment analysis.
  • Integrated with Google Cloud Storage.


AutoML Speech allows you to convert audio to text, through an API.

  • It returns text results in real time.
  • Context-aware recognition.
  • Automatic speech recognition.
  • Global vocabulary: many languages supported.
  • Streaming recognition.
  • Word hints.
  • Real-time or prerecorded audio support.
  • Noise robustness.
  • Inappropriate content filtering.

Natural Language AI

Derive insights from unstrucutred text using Google machine learning.

  • Train high-quality machine learning custom models without code, with AutoML.
  • Extract key document enitties that mattern.
  • Multimedia and multilingual support.
  • Content Classification Relationship Graphs.
  • Syntax Analysis.
  • Entity Recognition.
  • Sentiment Analysis.
  • Content classification.
  • Multilanguage.

Translation AI

Make your content and apps multilingual with fast, dynamic machine translation.

  • It supports real-time translation.
  • AutoML Translation, Translation API and Media Translation API
  • Many languages supported.
  • Language detection.