Google introduces Cloud centered Video Intelligence API

In a bid to make machine learning a transformational tool for organizations, Google has unveiled a cloud-based Video Intelligence Application Programming Interface (API) that helps locate the exact frame and video content by providing information about entities (nouns such as dog, flower or verbs such as run, swim) inside the video content.

The Video Intelligence API, could even provide contextual understanding of when those entities appear; for example, searching for tiger would find all precise shots containing tigers across a video collection in Google Cloud Storage, quoted Chief Scientist Google Cloud (Artificial Intelligence and Machine Learning) Fei Fei Li at the recent Cloud Next conference.


These APIs could let customers build the next generation of applications that could see, hear and understand unstructured data, greatly expanding the use cases for machine learning. The novel Cloud Video Intelligence API, which is currently in private beta, will use deep-learning models, built using frameworks like TensorFlow, to be applied on large media platforms such as YouTube. Very little metadata is available for Media organizations, which rely mostly on years of archived content alone. These could now use Video Intelligence API to understand key entities enabling them to find new ways to re-purpose their existing video.

The API quickly interprets videos stored on Google Cloud Storage with video, shot and frame-level context. However the early focus of the Silicon Valley giant as of now remains to offer an enterprise solution for large organizations and consumer technology companies.

Source: #-Link-Snipped-#