2 papers
cs.CV2026
ELASTIC: Efficient Once For All Iterative Search for Object Detection on Microcontrollers
Tony Tran, Qin Lin, Bin Hu
Deploying high-performance object detectors on TinyML platforms poses significant challenges due to tight hardware constraints and the modular complexity of modern detection pipeli…
cs.LG2025
SOLAR: Switchable Output Layer for Accuracy and Robustness in Once-for-All Training
Shaharyar Ahmed Khan Tareen, Lei Fan, Xiaojing Yuan +2
Once-for-All (OFA) training enables a single super-net to generate multiple sub-nets tailored to diverse deployment scenarios, supporting flexible trade-offs among accuracy, robust…