EfficientNet-lite

EfficientNet-lite are a set of mobile/IoT friendly image classification models. Notably, while EfficientNet-EdgeTPU that is specialized for Coral EdgeTPU, these EfficientNet-lite models run well on all mobile CPU/GPU/EdgeTPU.

Due to the requirements from edge devices, we mainly made the following changes based on the original EfficientNets.

Remove squeeze-and-excite (SE): SE are not well supported for some mobile accelerators.
Replace all swish with RELU6: for easier post-quantization.
Fix the stem and head while scaling models up: for keeping models small and fast.

Here are the checkpoints, and their accurracy, params, flops, and Pixel4's CPU/GPU/EdgeTPU latency.

Model	params	MAdds	FP32 accuracy	FP32 CPU latency	FP32 GPU latency	FP16 GPU latency	INT8 accuracy	INT8 CPU latency	INT8 TPU latency
efficientnet-lite0 ckpt	4.7M	407M	75.1%	12ms	9.0ms	6.0ms	74.4%	6.5ms	3.8ms
efficientnet-lite1 ckpt	5.4M	631M	76.7%	18ms	12ms	8.0ms	75.9%	9.1ms	5.4ms
efficientnet-lite2 ckpt	6.1M	899M	77.6%	26ms	16ms	10ms	77.0%	12ms	7.9ms
efficientnet-lite3 ckpt	8.2M	1.44B	79.8%	41ms	23ms	14ms	79.0%	18ms	9.7ms
efficientnet-lite4 ckpt	13.0M	2.64B	81.5%	76ms	36ms	21ms	80.2%	30ms	-

CPU/GPU/TPU latency are measured on Pixel4, with batch size 1 and 4 CPU threads. FP16 GPU latency is measured with default latency, while FP32 GPU latency is measured with additional option --gpu_precision_loss_allowed=false.
Each checkpoint all contains FP tflite and post-training quantized INT8 tflite files. If you use these models or checkpoints, you can cite this efficientnet paper.

Comparing with MobileNetV2, ResNet-50, and Inception-V4, our models have better trade-offs between accuracy and size/latency. The following two figures show the comparison among quantized versions of these models. The latency numbers are obtained on a Pixel 4 with 4 CPU threads.

As Tensorflow Lite also provides GPU acceleration for float models, the following shows the latency comparison among float versions of these models. Again, the latency numbers are obtained on a Pixel 4.

A quick way to use these checkpoints is to run:

$ export MODEL=efficientnet-lite0
$ wget https://storage.googleapis.com/cloud-tpu-checkpoints/efficientnet/${MODEL}.tar.gz
$ tar zxf ${MODEL}.tar.gz
$ wget https://upload.wikimedia.org/wikipedia/commons/f/fe/Giant_Panda_in_Beijing_Zoo_1.JPG -O panda.jpg
$ wget https://storage.googleapis.com/cloud-tpu-checkpoints/efficientnet/eval_data/labels_map.txt
$ python eval_ckpt_main.py --model_name=$MODEL --ckpt_dir=$MODEL --example_img=panda.jpg --labels_map_file=labels_map.txt

TFLite models can be evaluated using this tool.

Training EfficientNet-lite on Cloud TPUs

Please refer to our tutorial: https://cloud.google.com/tpu/docs/tutorials/efficientnet

Post-training quantization

$ export MODEL=efficientnet-lite0
$ wget https://storage.googleapis.com/cloud-tpu-checkpoints/efficientnet/${MODEL}.tar.gz
$ tar zxf ${MODEL}.tar.gz
$ python export_model.py --model_name=$MODEL --ckpt_dir=$MODEL --data_dir=/path/to/representative_dataset/ --output_tflite=${MODEL}_quant.tflite

To produce a float model that bypasses the post-training quantization:

$ python export_model.py --model_name=$MODEL --ckpt_dir=$MODEL --output_tflite=${MODEL}_float.tflite --quantize=False

The export_model.py script can also be used to export a tensorflow saved_model from a training checkpoint:

$ python export_model.py --model_name=$MODEL --ckpt_dir=/path/to/model-ckpt/ --output_saved_model_dir=/path/to/output_saved_model/ --output_tflite=${MODEL}_float.tflite --quantize=False