Skip to content

This GitHub repository aims to provide comprehensive overview of the evolution of GoogLeNet, a popular Convolutional Neural Network architecture developed by Researchers at Google for image classification tasks.

License

Notifications You must be signed in to change notification settings

cx-olquinjica/InceptionNet-Variations

Repository files navigation

InceptionNet-Variations

This GitHub repository aims to provide comprehensive overview of the evolution of GoogLeNet, a popular Convolutional Neural Network architecture developed by Researchers at Google for image classification tasks. The repository will showcase the different iterations of the GoogLeNet architecture, from the first paper Szegedy et al. in 2015, which introduced the Inception module, to the later versions of the model proposed by Szegedy et al. in 2016 and 2017. Each iteration of the model is implemented using popular deep learning frameworks such as TensorFlow and PyTorch, and is trained and evaluated on standard images classification datasets such as CIFAR-10 and ImageNet. The repository provides code and documentation for each implementation, making it valuable resource for researchers and practitioners interested in the history and development of the GoogLeNet architecture.

They include the following:

Run

Using config files

Modify the configurations in .json config files, then run:

python train.py --config config.json

Resuming from checkpoints

You can resume from a previously saved checkpoint by:

python train.py --resume path/to/checkpoint

Using Multiple GPU

You can enable multi-GPU training by setting n_gpu argument of the config file to larger number. If configured to use smaller number of gpu than available, first n devices will be used by default. Specify indices of available GPUs by cuda environmental variable.

python train.py --device 2,3 -c config.json

This is equivalent to

CUDA_VISIBLE_DEVICES=2,3 python train.py -c config.py

Citations

@misc{ioffe2015batch,
      title={Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift}, 
      author={Sergey Ioffe and Christian Szegedy},
      year={2015},
      eprint={1502.03167},
      archivePrefix={arXiv},
      primaryClass={cs.LG}
}
@misc{szegedy2015rethinking,
      title={Rethinking the Inception Architecture for Computer Vision}, 
      author={Christian Szegedy and Vincent Vanhoucke and Sergey Ioffe and Jonathon Shlens and Zbigniew Wojna},
      year={2015},
      eprint={1512.00567},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}
@misc{szegedy2016inceptionv4,
      title={Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning}, 
      author={Christian Szegedy and Sergey Ioffe and Vincent Vanhoucke and Alex Alemi},
      year={2016},
      eprint={1602.07261},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}

About

This GitHub repository aims to provide comprehensive overview of the evolution of GoogLeNet, a popular Convolutional Neural Network architecture developed by Researchers at Google for image classification tasks.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published