Smart Multimedia Information Retrieval and the SMIRA Platform can provide solutions for various applications in different areas. Based on the powerful integration, extraction, fusion, and conversion of multimedia features, a lot of use cases can be built to support existing or future applications. On this page, a selection of some application areas is given, where SMIRA can deliver effective, efficient and flexible solutions.
Automotive
Modern Vehicles produce a stream of thousands of different Multimedia Features per second. Sensors, cameras, surveillance systems, telemetrics, driver interactions, audio signals (like sirens or horns) – they all can be regarded as multimedia feature with different relevance. The integration of such features can be employed to model certain states (like security alerts). The SMIRA Platform can integrate these features, calculate relevance, fuse information, and provide realtime insight.
Broadcast
Broadcast is one of the core and original domains, where the integration of Multimedia is required. The SMIRA Platform is compatible to all major broadcast protocols, like MOS, BXF, MXF, MPEG7 and supports the integration and feature extraction of all media types including images, text-based annotations, SQL databases, and XML-sidecar media descriptors. Furthermore, export of feature graphs, indexing structures, or textual and human-understandable feature representations into various formats, is possible. This enables – amongst others – the automated generation of archiving information, audio and speech detection, and a highly effective querying of media archives.
Social Media
In the area of Social Media, particularly the integration and fusion of Multimedia information, is important. Relevant information of a Social Media Post is not just the image or video – also related texts, comments, the combination of image, video, description, location, and feedback has to be represented as Multimedia feature. Here, the calculation of relevance and the weighted fusion of features into a “real” description and annotation of a Social Media post, is required.
Medical Applications
The area of medical applications highly utilizes Multimedia objects. One major challenge in this area is the integration of textual information (e.g., doctor’s letters), image and video technologies, surveillance monitors, and value based features (like blood values). Furthermore, the historical information of features and their development, is required for diagnosis. Smart Multimedia Information Retrieval provides calculation models for the exact detection of changed medical features and correlations to other multimedia sources. Furthermore, the explainability functions of the Platform can deliver human understandable texts describing technical features for doctors, nurses, patients and family members.
Media Production and Archiving
The SMIRA Platform can provide solutions for the automated extraction of features from video, audio, and text content. Production Companies can employ SMIRA to detect objects from videos, extract audio information (title, artist, composer) from a video’s audio track, apply speech-to-text algorithms for news or talk shows, and calculate an integrated metadata record for a video or live-stream automatically. Due to many adapters and integrations with standard tools, this record can be used in further process steps.
Private Producers and Consumers
Of course, Smart Multimedia Information Retrieval can be employed for any end-user application based on multimedia objects and will deliver fast and effective image or video search, semantic filters and querying, or human-understandable explanations for desktop or mobile applications.