1.1.0 Data Possibility to process offline data, connect cameras and analyse live processing streams
1.2.0 Distributed Possibility to locally analyse faces in camera streams on autonomous,
live analysis remote points. Immediate transfer of sightings and results to the center for evaluation; transfer tolerant against network outages (local buffering of results and transfer at active network connection)
1.3.0 Dependency Pure software solution, installable by end-user on commercially available
2.1.0 Interfaces Possibility to connect external systems and to call configuration, import and search functions via http-REST integration API
2.1.1. Interface data Possibility to automatically import data (video and images), e.g. triggered by import external systems
2.1.2 Interface Possibility to register and parametrise of cameras via the integration API channel management
2.1.3 Interface Possibility to automatically query recognised faces (including scores and events and thumbnails) in offline imported or live analysed data. Possibility to forward alerts alerts to external systems (when persons enrolled to a blacklist are spotted on registered cameras)
2.1.4 Interface Possibility to automatically create, manage, enhance and delete identities identity via the integration API
Unique Selling Points of Videmo 360 Page 1
management
2.1.5 Interface Possibility to automatically carry out search queries (with certain filter search queries criteria, identities, age and gender) by external systems via the integration API. Return of results as sorted lists (including scores and thumbnails) for external viewing and further processing
2.2.0 Integration Integration with video management system from manufacturer Dallmeier
2.3.0 Integration Integration with video management system from manufacturer Geutebrück
3. Connection of cameras ID Name Description
3.1.0 Camera types Connection to different camera types (USB cameras, IP cameras with mjpeg over http, rtsp, machine vision cameras for analysis free from compression)
3.2.0 Proprietary Connection to Dallmeier Netbox and comfortable registering of connected
systems cameras there via the user interface
4. Import ID Name Description
4.1.0 Video formats Import of all established video formats that are supported by FFMPEG and HEIF
4.2.0 Batch import Possibility of to import up to 1Mio video files in a batch import
4.3.0 Media Unlimited import without copying data. Videmo accesses the folders directly Locations on the server
4.3.0 Proprietary Import of video material from manufacturer Dallmeier
formats
Unique Selling Points of Videmo 360 Page 2
5. Search and find ID Name Description
5.1.0 Search modes - Search existing video material or camera channel events for certain Identity identities based off a photo or video.
5.1.1 Search modes Search for persons based on facial attributes (such as gender, age, beard, –facial type/color of hair, etc.) in offline data and in sightings delivered from live attributes analysis
5.1.2 Search mode – Search for persons as a whole and person attributes such as clothing and person gender attributes
5.1.3 Search mode - Search for vehicles (cars, trucks, bicycle etc.) and luggage (suitcase, objects handbags etc.)
5.1.4 Search mode - Search for License plates either by entering the license plate or by ALPR displaying all license plates in the data
5.2.0 Identification Display of alerts (recognitions) of persons registered in a blacklist when
showing up on registered cameras
5.3.0 Cross-media Offline search for faces in single images as well as in video data by using search identity models created and managed cross-medial
5.4.0 Detection Display of all detected faces (filtered for various criteria), as thumbnails in graphical user interface
5.5.0 Iterative Iterative search with gradually refined face models and additional filters search mode (age, gender, time, media) for more precise results
5.6.0 Parallel search Possibility to search in parallel for multiple identities
5.6.1 Parallel search Different searches in multiple tabs
5.7.0 Live sightings Fast sighting of faces detected in live video streams
5.8.0 Tracking and Tracking of faces in video and video-based matching with face recognition matching (instead of matching single templates per sighting)
Unique Selling Points of Videmo 360 Page 3
6. Management of data and identities ID Name Description
6.1.0 Methods Possibility to train identities by uploading face images as well as by simple enrollment drag-and-drop of sighted target persons directly from analysed image and video material
6.2.0 Media Management of imported video data, imported image collections (image management databases) as well as remote connected cameras via the user interface
6.2.1 Media Media Locations mirror the folder and data structure on the server and are management synchronized on command
6.3.0 Enhancement Possibility to enhance interactively identity models with new image material identity by simple drag-and-drop models
6.4.0 Synchroni- Automatic synchronisation of central blacklists with distributed viewing
sation of points for local analysis and alerting identities
6.5.0 Case Possibility to separate cases via workspaces, tiered role management Management
6.6.0 Identity Identities can be grouped and tagged for easy organization management
6.6.1 Identity Mass Enrollment allows for Batch-Enrollment of Identities with automated management naming pattern
7. Unique selling points company
ID Name Description
7.1.0 Method Face recognition algorithm is developed in-house
7.2.0 Cooperation Cooperation with the Karlsruhe Institute of Technology, a driver of
innovation in the research about face recognition
7.3.0 Cooperation Long-time research and development cooperation with police authorities
7.4.0 Cooperation Participation in German and EU research projects in the field of face recognition for security applications
