Infosys media platform capitalizes on current & innovative technologies to facilitate creative collaboration, enable a production on an industrial scale, and monetize customer relationships. The unifying framework integrates the various ecosystems and provides a common platform to deliver services and technology solutions for the media content value chain.
The intelligently woven media and metadata management framework uses smart workflows and incorporates machine learning and artificial intelligence to create, manage, moderate content metadata, and orchestrate workflows across different business functions. Additionally, the platform delivers the benefits of productivity, scale, and agility via the cloud and streamlines collaboration between the plethora of partners and technology solutions across the supply chain.
The Platform consists of various modules to serve different business functions.The Metadata Management module orchestrates the accumulation of asset level and descriptive time-based metadata from production to delivery.
The Curation Module provides the master workflow and ingests content from internal archives and multiple sources and includes functions like automated QC, Editing, Review, and Approval, and censor editing.
The Archive Digitization module uses AI/ML to create a composite index of frame level, time-coded metadata of recognized elements such as celebrities/known personalities, objects, brands, text, and images along with enabling intelligent Ad spot identification.
The Custom AI modeling capability ensures that newly introduced elements such as celebrities, brands, etc. can be continuously trained and recognized.
The Localization Module enables Automatic generation of close captions and subtitles in multiple languages while enabling collaboration across multiple locations and vendors present around the globe.
The Platform’s Digital Distribution Module has standard workflow models and a state-of-the-art dashboard to enable global distribution to digital platforms at scale.
Leveraging AI models to process video files to generate time coded metadata for post-production and distribution processes. Infosys Media Platform can recognize multiple metadata attributes like celebrities, brands, music, shot changes, text visible on screen and any other objects visible in the video frames. Once the metadata is enriched, this can be used for compliance checks, censor editing or in providing real time metadata driven contextual and dynamic ad-placements.
Infosys Media Platform comes with capability to process the audio (dialogs from a video, lyrics from a song, speech from a podcast) and convert the same into closed captions and subtitles. The platform also has built in capabilities to recognize the language on its own and further translate it on demand into 30+ languages without taking any additional processing time. This will help a content provider to distribute content globally with a significant reduction in time to market.
Using the video content and speech detection capabilities, Infosys Media Platform can recognize presence of mature content (profanity, violence, gore etc.). The information is provided at the shot level which enables a censor editor to easily review and clip content. Alternately the same information can be used by video players to jump content at run time to provide “demography appropriate” viewing experience.
Similar to the capabilities on Video/Music content, Infosys Media Platform has ability to identify various attributes of an image file. A number of image attribute driven actions can be performed using ability to recognize celebrities, brands, emotions, objects and context from a file. A good example could be automated advertising quality review process for publishing ads on an online platform. Or using OCR based text recognition to perform text analytics.
Infosys Media Platform comes with capability to manage end to end supply chain of digital metadata creation, updates, packaging and distribution. Driven by smart workflows, the platform can interface with any systems to collect, compare or update metadata; perform automated quality checks and package it in the format requested by third party content consumers.
Using natural language processing capabilities of the platform, you can review any string of text (dialogs, lyrics, conversations) to determine the context of the conversation as well as the sentiment. This in turn helps in digital content consumption analytics by adding the dimension of sentiment further improving the personalization and recommendation capabilities.
Unified media workflows – Connecting various parties in the supply chain for seamless exchange of information and automated update and delivery of metadata
AI-based metadata generation – To create metadata automatically at scale reducing human intervention and monetize archived content
Multi-language translations & subtitling – Creating closed captions and converting the same into multiple languages providing localization at scale globally
1. Time coded metadata attributes – Generating metadata at a shot level enabling real-time personalized actions and content moderation
2. Custom AI models – Providing the ability to train your own AI models for custom object detection catering to business-specific metadata requirements of the customers
Open source, cloud-based, modular & elastic offering –Bringing operational scalability, functional portability and cost flexibility to the customers
Fully automated processes bringing down cost and reducing manual intervention