Page 314 - AI for Good Innovate for Impact

AI for Good Innovate for Impact

• Participate in the development of standards and regulations related to AIGC content
security detection and protection, promoting the healthy development of the industry.
• Establish reference tools and simulation environments for AIGC content security detection
and protection to help users better understand and apply the platform.

3 Use Case Requirements

• REQ-01: It is critical that a high-quality AI security corpus pipeline based on cognitive
generation is both deployed, as it forms the foundation for training, evaluating, and
improving the system's ability to detect and respond to security risks in AI-generated
content.
• REQ-02: It is critical that the system is capable of continuously generating evaluation data
and training corpora.
• REQ-03: It is critical that algorithmic models are evaluated and protected automatically
to ensure the trustworthiness of AI systems in operational environments.

4 Sequence Diagram

• SRC (source): This node is the source of data that can be used as input to the ML pipeline.
• C (collector): This node is responsible for collecting data from one or more SRC nodes.
• PP (preprocessor): This node is responsible for cleaning data, aggregating data or
performing any other preprocessing needed.
• M (model): This is a machine learning model
• P (policy): This node enables the application of policies to the output of the model node.
• D (distributor): This node is responsible for identifying the SINK(s) and distributing the
output of the M node.
• SINK: This node is the target of the ML output on which it takes action.

5 References

[1] J. Kim, M. Hyun, I. Chung, and N. Kwak, “Feature Fusion for Online Mutual Knowledge
Distillation,” arXiv preprint:1904.09058*, Apr. 2019. [Online]. Available: https:// arxiv �org/
pdf/ 1904 �09058

278