InceptionV3: Rethinking the Inception Architecture for Computer Vision
InceptionV3 is an upgraded version of GoogLeNet. One of the most important improvements of V3 is Factorization, which decomposes 7x7 into two one-dimensional convolutions (1x7, 7x1), and 3x3 is the same (1x3, 3x1), such benefits, both It can accelerate the calculation (excess computing power can be used to deepen the network), and can split 1 conv into 2 convs, which further increases the network depth and increases the nonlinearity of the network. It is also worth noting that the network input from 224x224 has become 299x299, and 35x35/17x17/8x8 modules are designed more precisely. In addition, V3 also adds batch normalization, which makes the model converge more quickly, which plays a role in partial regularization and effectively reduces overfitting.[1]
Figure 1. Architecture of InceptionV3 [1]
Our reproduced model performance on ImageNet-1K is reported as follows.
| Model | Context | Top-1 (%) | Top-5 (%) | Params (M) | Recipe | Download |
|---|---|---|---|---|---|---|
| inception_v3 | D910x8-G | 79.11 | 94.40 | 27.20 | yaml | weights |
- Context: Training context denoted as {device}x{pieces}-{MS mode}, where mindspore mode can be G - graph mode or F - pynative mode with ms function. For example, D910x8-G is for training on 8 pieces of Ascend 910 NPU using graph mode.
- Top-1 and Top-5: Accuracy reported on the validation set of ImageNet-1K.
Please refer to the installation instruction in MindCV.
Please download the ImageNet-1K dataset for model training and validation.
- Distributed Training
It is easy to reproduce the reported results with the pre-defined training recipe. For distributed training on multiple Ascend 910 devices, please run
# distributed training on multiple GPU/Ascend devices
mpirun -n 8 python train.py --config configs/inceptionv3/inception_v3_ascend.yaml --data_dir /path/to/imagenetIf the script is executed by the root user, the
--allow-run-as-rootparameter must be added tompirun.
Similarly, you can train the model on multiple GPU devices with the above mpirun command.
For detailed illustration of all hyper-parameters, please refer to config.py.
Note: As the global batch size (batch_size x num_devices) is an important hyper-parameter, it is recommended to keep the global batch size unchanged for reproduction or adjust the learning rate linearly to a new global batch size.
- Standalone Training
If you want to train or finetune the model on a smaller dataset without distributed training, please run:
# standalone training on a CPU/GPU/Ascend device
python train.py --config configs/inceptionv3/inception_v3_ascend.yaml --data_dir /path/to/dataset --distribute FalseTo validate the accuracy of the trained model, you can use validate.py and parse the checkpoint path with --ckpt_path.
python validate.py -c configs/inceptionv3/inception_v3_ascend.yaml --data_dir /path/to/imagenet --ckpt_path /path/to/ckptPlease refer to the deployment tutorial in MindCV.
[1] Szegedy C, Vanhoucke V, Ioffe S, et al. Rethinking the inception architecture for computer vision[C]//Proceedings of the IEEE conference on computer vision and pattern recognition. 2016: 2818-2826.