WRSWorld Research StateContribute
WORLD RESEARCH STATE

MnasNet platform-aware mobile architecture search

Contribution grade

Hardware-aware neural architecture search · Human-designed research method or resource

Reference score130,000Provisional editorial estimate of this exact contribution; its grade is assessed separately.

What changed

Searches a factorized hierarchical mobile CNN space with a reward balancing measured target-phone latency and classification accuracy.

Scientific scope

Searches a factorized hierarchical mobile CNN space with a reward balancing measured target-phone latency and classification accuracy. Automated search selected mobile network designs using accuracy and measured latency; researchers supplied the search space, devices and evaluation. Original v1 uses proxy-task search and transfer, not a claim that every candidate is fully trained on ImageNet. A weighted product approximates nearby tradeoffs, not an exhaustive certified Pareto frontier; latency is specific to measured hardware.

Evidence

Automated search selected mobile network designs using accuracy and measured latency; researchers supplied the search space, devices and evaluation. WRS has not independently rerun these experiments. Automated search selected mobile network designs using accuracy and measured latency; researchers supplied the search space, devices and evaluation. Original v1 uses proxy-task search and transfer, not a claim that every candidate is fully trained on ImageNet. A weighted product approximates nearby tradeoffs, not an exhaustive certified Pareto frontier; latency is specific to measured hardware.

Contributors

Provisional WRS assessment, edition 2026-10-16.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.

Full scientific record · Browse contribution grades · Request a correction