SDVI is an Emmy® Award-winning supplier of cloud-based media supply chain technology that empowers organizations to optimize content ingest, processing, packaging, and distribution operations. The company’s Rally media supply chain platform helps organizations create a scalable and responsive infrastructure that provides true business agility, operational efficiency, and process intelligence.
DeepVA products used
Rally media supply chain platform
The Rally platform optimizes media supply chains by deploying the applications and the cloud resources they need for each step in a supply chain. These applications are activated, monitored, and then shut down as needed, enabling operators to choose the right application for every job and only paying for what is used. Using Rally’s orchestration and decision engine, automated supply chains are built using tightly integrated 3rd party tools for content analysis, transformation, QC, audio leveling, metadata enrichment, and more.
The integration of DeepVA tools with Rally means that operators now can access DeepVA’s content analysis capabilities as part of Rally-managed supply chains. Rally will normalize and associate the metadata generated by DeepVA with each asset so that operators can use the enriched metadata to facilitate better search results and monetization efforts. You can find more information here.
Benefits for the user
DeepVA features integrated into Rally platform
Face Recognition detects and identifies the faces of public figures in a variety of categories such as politics, sports, business, and entertainment.
Often important information is already contained in the image, our text recognition recognizes it and makes it usable.
Landmark Recognition identifies all important sights, architectural structures and natural monuments across Europe and North America.
Visual concepts recognition
Object and scene recognition detects and labels various objects and scenes, from general to more specific ones.
Face Attributes recognizes emotions and facial characteristics such as "beard", "eyes closed" or "glasses" of all persons appearing in pictures or videos.
Diversity Analysis offers the possibility to determine the percentage of gender and age occurrence in images or videos.
Using QR codes, important information and links can be integrated directly in the footage, our QR code detection then makes them readable again.
Our logo recognition helps you to build your brand by monitoring the presence in media, as well as the use of the right logo.
Often the spoken word alone is not enough: transcription is needed. Our Speech to Text function automates this process.
Aspect ratio detection
Detect the format of your images and videos and intelligently crop them to suit your needs.
Shot boundary detection
Precisely segment your video to skip intros and recaps, and detect title sequences, voice-overs, or break bumpers.
Identify a variety of personalities by their voice using real-time speaker identification.
In just a few seconds, new faces are trained using our face training technology and are ready for use in Recognition Services.
With just a few images, our Landmark Training allows you to efficiently train landmark buildings.
Extend the logo recognition with your own logos and train your AI models with just a few clicks.
Your speakers can be trained easily and intuitively with very little training data and no negative examples.
Using Face Index, each person in video and image material can be assigned a number, allowing them to be translated into the metadata afterwards.
Knowledge Graph recognizes the relationships between people or objects, represents them visually, and thus helps to stay one step ahead of the competition.
Face Dataset creation
Using images of the faces and text information (e.g., from belly bands), Face Dataset Creation automatically creates entire datasets, without any manual labor.
Face Evaluation analyses entire datasets and gives users feedback on the quality of their training data. This improves the quality of the dataset and makes face recognition more accurate.
A few pictures of the landmark are enough to create a dataset and make it possible to find the footage that contains that particular place or landmark.