Need to support several models

Context

To complete the benchmark, several models need to be added to the CPP export :

Task	Dataset	Model	CPP fp32	CPP i8	Bench
KeyWord spotting	Google Speech Commands	DS-CNN	❌	❌	❌
Image classification	ImageNet (224x224)	MobileNet-v1	✔️	🔨	❌
Visual wake words	Visual Wake Words Dataset	MobileNet-v1 0.25x	✔️	✔️	❌
Object detection	COCO (300x300)	SSD-MobileNet-v1	❌	❌	❌
Object detection	MS-COCO 2017	SSD-MobileNet-v2	❌	❌	❌
Image classification	CIFAR10	ResNet-8	✔️	✔️	❌
Image classification	ImageNet (224x224)	Resnet50-v1.5	✔️	✔️	❌
Anomaly detection	ToyADMOS	Deep AutoEncoder	❌	❌	❌
Image segmentation	KITS 2019 (602x512x512)	3D UNET	❌	❌	❌
Language processing	SQUAD 1.1	MobileBert	❌	❌	❌