SqueezeNet

Use case : Image classification

Model description

SqueezeNet is a pioneering compact architecture that achieves AlexNet-level accuracy with 50x fewer parameters. It introduced the "Fire module" combining squeeze and expand operations.

The architecture features Fire Modules with squeeze (1x1) followed by expand (1x1 + 3x3) layers, employing delayed downsampling to maintain larger activation maps longer. It uses no fully connected layers, relying on global average pooling, resulting in an extremely compact model (<0.5MB original size).

SqueezeNet is ideal for extremely constrained deployment scenarios, model compression research, and applications where model size is critical.

(source: https://arxiv.org/abs/1602.07360)

The model is quantized to int8 using ONNX Runtime and exported for efficient deployment.

Network information

Network inputs / outputs

For an image resolution of NxM and P classes

Recommended platforms

Performances

Metrics

• Measures are done with default STEdgeAI Core configuration with enabled input / output allocated option.
• All the models are trained from scratch on Imagenet dataset

Reference NPU memory footprint on Imagenet dataset (see Accuracy for details on dataset)

Reference NPU inference time on Imagenet dataset (see Accuracy for details on dataset)

Accuracy with Imagenet dataset

Retraining and Integration in a simple example:

Please refer to the stm32ai-modelzoo-services GitHub here

References

[1] - Dataset: Imagenet (ILSVRC 2012) — https://www.image-net.org/

[2] - Model: SqueezeNet — https://github.com/DeepScale/SqueezeNet