Graduates of this program will possess the technical proficiency to design, optimize, and deploy efficient AI solutions on resource-constrained edge devices.
- Optimize machine learning models for TinyML applications by applying advanced compression techniques to reduce memory footprint and computational overhead.
- Design low-latency, real-time inference pipelines that ensure reliable data processing and immediate decision-making capabilities in edge environments.
- Integrate specialized hardware accelerators, such as NPUs and GPUs, to enhance the computational throughput and energy efficiency of edge AI systems.
- Implement model quantisation and pruning strategies to significantly reduce model size while maintaining acceptable accuracy levels for deployment.
- Deploy federated learning frameworks at the edge to enable collaborative model training across distributed devices while preserving data privacy and security.
- Develop robust edge computer vision systems capable of performing real-time image and video analysis with minimal power consumption.
- Evaluate AI models for ethical compliance and bias mitigation to ensure fair and responsible deployment of edge intelligence in diverse contexts.
- Benchmark the performance of edge AI solutions using standardized metrics for latency, throughput, and power efficiency to validate system readiness.