Optimize image filters on device
Company: Apple
Role: Machine Learning Engineer
Category: ML System Design
Difficulty: medium
Interview Round: Onsite
You are shipping an image-filter feature that must run entirely on a mobile device. Users expect preview latency below 30 ms on common phones, memory overhead to stay low, and battery impact to be minimal. Describe how you would optimize the filter pipeline while preserving acceptable visual quality.
Your discussion should cover algorithm choice, numerical precision, CPU versus GPU or accelerator usage, memory layout, tiling or streaming, thermal constraints, and how you would measure quality after optimization.
Overview: This question evaluates competency in on-device machine learning and performance engineering, covering algorithm choice, numerical precision trade-offs, hardware utilization (CPU/GPU/accelerator), memory layout, tiling/streaming, and trade-offs among latency, memory overhead, and battery impact.