[NeurIPS 2020] MCUNet: Tiny Deep Learning on IoT Devices; [NeurIPS 2021] MCUNetV2: Memory-Efficient Patch-based Inference for Tiny Deep Learning; [NeurIPS 2022] MCUNetV3: On-Device Training Under 256KB Memory
The ImageNet variants of mcunet are producing inconsistent predictions, likely due to pre-training on ImageNet classes rather than person prediction. As there's no .cpp code in the official repo for detecting these objects, could we obtain code to address this issue?
The ImageNet variants of mcunet are producing inconsistent predictions, likely due to pre-training on ImageNet classes rather than person prediction. As there's no
.cpp
code in the official repo for detecting these objects, could we obtain code to address this issue?