Page 110 - AI for Good-Innovate for Impact Final Report 2024

AI for Good-Innovate for Impact

23�3� Use case requirements

ITU-T Supplement Y.71 ITU-T Y.3000 series – Use Cases for Safety Monitoring
• GAI-VA-SM-UC23-REQ-001: The GAI-VA-SM system must be capable of operating
within the domain of safety monitoring, specifically focusing on improving worker safety
in high-risk industries like manufacturing. The system should be designed to handle a
high volume of video data and process it in real-time.
• GAIVASM-UC23- REQ-002: The system must be capable of addressing the significant
lead time and effort required to create models for safety monitoring in enterprises. It
should leverage advanced machine learning techniques to automate and expedite the
model creation process.
• GAIVASM-UC23-REQ-003: The system should utilize fine-tuned models for video
analytics at the edge with real-time monitoring. These models should be trained using
visual prompting and images from existing CCTV cameras. The system should also be
capable of handling a variety of video formats and resolutions.
• GAIVASM-UC23-REQ-004: The system should employ zero-shot object detection and
visual prompting techniques. It should be capable of identifying safety violations even in
scenarios that it has not been explicitly trained on.
• GAIVASM-UC23-REQ-005: The system should be designed to be tested in dedicated
testbed environments resembling real-world industrial settings and be capable of
being deployed in pilot projects in real-world manufacturing, construction, and logistics
environments. The system should also be robust enough to handle varying lighting
conditions and different camera angles.

23�4� Sequential diagram

23�5� References

[1] Securade.ai - Safety, powered by AI

94