Thank you for sending your enquiry! One of our team members will contact you shortly.
Thank you for sending your booking! One of our team members will contact you shortly.
Duration 21 hours
Course Outline
Introduction to TinyML Security
- Security hurdles in resource-limited machine learning systems
- Threat modeling for TinyML implementations
- Risk classifications for embedded AI applications
Privacy in Edge AI Data Handling
- Privacy implications of on-device data processing
- Strategies to reduce data exposure and transmission
- Methods for decentralized data management
Adversarial Attacks on TinyML Models
- Threats involving model evasion and poisoning
- Input manipulation targeting embedded sensors
- Evaluating vulnerabilities within constrained environments
Enhancing Security in Embedded ML
- Firmware and hardware protection layers
- Access control protocols and secure boot processes
- Best practices for protecting inference pipelines
Privacy-Centric TinyML Techniques
- Quantization and model design strategies with a privacy focus
- On-device anonymization methods
- Lightweight encryption and secure computation approaches
Secure Deployment and Upkeep
- Secure provisioning of TinyML devices
- Over-the-air updates and patch management strategies
- Edge-level monitoring and incident response
Testing and Verification of Secure TinyML Systems
- Frameworks for security and privacy testing
- Simulating real-world attack vectors
- Validation and regulatory compliance considerations
Case Studies and Practical Scenarios
- Security lapses in edge AI ecosystems
- Architecting robust TinyML solutions
- Assessing the balance between performance and protection
Summary and Path Forward
Requirements
- Familiarity with embedded system architectures
- Practical experience with machine learning workflows
- Foundational knowledge of cybersecurity principles
Target Audience
- Security analysts
- AI developers
- Embedded engineers
Testimonials (1)
The exercises
