RT-CAS : real-time context-aware scheduling for mixed deep learning tasks with a case study in colonoscopy
Rawit Lertluksanaporn
Modern real-time deep learning (DL) systems increasingly execute multiple heterogeneous inference tasks concurrently rather than a single isolated task. This is especially true in safety-critical applications such as real-time colonoscopy assistance, where inference tasks differ in execution patterns, urgency levels, a...
- Office of Academic Resources, Chulalongkorn U...
- DOI: 10.58837/chula.the.2025.871