Internship / Master’s Thesis – Embedded AI
Core
Implement hardware-dependent knowledge distillation for a compact multispectral object-detection student and run it on M.2 AI accelerators.
Role type
Research intern / Master's thesis candidate (Embedded AI)
Builds
Compact multispectral object-detection models optimized for edge deployment
Domain
Space technology / Edge AI / Multispectral Earth observation
Deliverable
production ML models
Required skills
knowledge distillation, model quantization, model compilation, edge AI hardware deployment, multispectral data processing
Preferred skills
resource-constrained system design, toolchain evaluation
Technologies
Axelera Embedded 110m, MemryX Cascade 100, M.2 AI accelerators, DPUs
Responsibilities
Train a compact student model using Hinton-style knowledge distillation; research and implement alternative distillation methods; compile and run the student on commercial M.2 accelerators; compare accuracy and throughput; assess toolchain usability; recommend optimal hardware.
Seniority
Intern / Master's Thesis
