Google Cloud Platform’s machine learning tools allow media and entertainment companies to make the most out of their data and redefine their business strategies to resonate with today’s digital consumers.
While the media and entertainment industry as a whole is projected to see modest but steady growth through 2021, traditional market segments are in decline as audiences find and consume content far differently than they did in the past. Channel flippers are giving way to cord cutters, and streaming services killed the video store. Today’s digital consumers want specific content delivered to their devices, they want to consume it on their own schedule, and they expect a seamless, world-class user experience.
M&E companies have turned to machine learning to better understand their target audiences so that they can deliver the content and the immersive, personalized experiences they demand. ML algorithms are being deployed to help design trailers and other advertisements, provide consumers with personalized content recommendations, and even streamline the production process.
As the undisputed leader in AI and ML, the Google Cloud Platform allows enterprises to harness Google’s decades of research into AI/ML and deploy advanced ML models without the need for a team of data scientists. With a wide selection of pre-trained ML models and easy-to-use tools that allow developers to build and train their own, GCP advances Google’s mission to make AI and ML easy, fast, and useful for all developers and enterprises.
Deploy pre-trained ML models with GCP machine learning APIs
GCP’s pre-trained, optimized ML models are great when an enterprise wants to get started with a specific use case quickly. Available APIs include:
- Google Cloud Video Intelligence, a REST API that allows developers to access Google’s video analysis technology to annotate entire videos, segments, shots, or frames with contextual information. Possible use cases include detection of labels, explicit content, and shot changes, as well as regionalization features.
- Google Cloud Vision API, which enables applications to access vision detection features. Possible use cases include detecting labels, faces, logos, and landmarks, performing optical character recognition (OCR), and tagging explicit content within images.
- Google Cloud Speech API, which supports over 110 languages and variants, allows applications to integrate speech recognition technologies for purposes of subtitles, closed-captioning, and other use cases where audio must be transcribed to text.
- Google Natural Language API, which enables applications to perform sentiment analysis, entity analysis, entity-sentiment analysis, content classification, and syntax analysis. Use cases include sentiment analysis of online film and TV reviews.
- Dialogflow Enterprise Edition, which allows users to build chatbots and other interactive, natural-sounding conversational interfaces for websites, mobile applications, popular messaging platforms, and IoT devices. Dialogflow can analyze multiple types of input, including text and audio, and can respond to consumers using text or synthetic speech.
Easily build & train custom ML models with Google Cloud AutoML
For use cases that fall outside the pre-trained APIs, there is Cloud AutoML, which is currently in beta. Using a graphical interface, developers without extensive machine learning knowledge can build and train custom ML models for specific business needs.
- AutoML Tables is a no-code solution that allows anyone, whether a data scientist, developer, or analyst, to build and deploy ML models on structured tabular datasets and incorporate them into wider applications. Models can be designed and deployed within days, as opposed to weeks. Fox Sports used AutoML Tables to build a model that successfully engaged Australian cricket fans by using historical data to predict when wickets would fall five minutes before a pitch.
- AutoML Vision, which was recently updated to support functionality at the edge, builds, trains, and optimizes models for use in applications in the cloud or at the edge to derive insights from images.
- AutoML Video Intelligence is a video analysis and annotation solution that allows developers to use custom labels. Use cases include creating highlight reels or automatically removing commercials.
- AutoML Natural Language uses machine learning to reveal the structure and meaning of text to glean more insight out of social media sentiment and customer conversations.
Much more to come through the new AI Platform
Google’s new AI Platform, announced at Next ‘19, is being billed as an end-to-end development platform for teams to build, run, and manage their ML projects through a shared interface. Since the AI Platform supports Kubeflow, Google’s open-source platform, it lets users build portable ML pipelines that can be run in the cloud or on-prem with very minimal code changes.