Cloud VMS provides a variety of pre-processing functions and interfaces for integration with AI and making AI processing cost efficient.
Filtering and sampling
The operations are performed in 3 steps:
Filtering based on camera events (motion, etc.).
Sampling and preparation of data in the most efficient format.
Interfacing with AI.
As the result of the first two steps, VXG significantly reduces the load on AI engines and the bandwidth, making AI operations cost-efficient.
Processing modes
VXG extracts data and makes it available in the following formats:
Timer-based images.
Event-based images.
Continuous video.
Event-based video.
Supported AI services
As the last step, VXG interfaces the data with AI engines. VXG is integrated with a variety of AI engines including:
Amazon Rekognition
Google Vision AI
Azure Cognitive Services
YOLO (hosted by VXG)
The 3-step approach results in huge cost and bandwidth savings.
AI features
For the full list of AI features please visit the corresponding AI services websites:
Amazon Rekognition:
For images: https://aws.amazon.com/rekognition/image-features/
For video: https://aws.amazon.com/rekognition/video-features/
Azure Cognitive Services:
https://azure.microsoft.com/en-us/services/cognitive-services/computer-vision/#features
Google Vision API:
https://cloud.google.com/vision#section-11
YOLO (open-source service hosted by VXG):