Image Classification Model in PyTorch for Indoor Scene Recognition.
This is a Google Colab Notebook. The notebook can also be found here : https://colab.research.google.com/drive/1ALGzKj3cafL0MPUNtbUNXzRc8S_C_lqM
The MIT-67 Indoor Scene Recognition Dataset is used here. The dataset has 15620 images in total distributed amongst 67 classes consiting of airport,trainstation,kitchen,library,etc. Link: http://web.mit.edu/torralba/www/indoor.html
The model used is Resnext101_32x16d. Link: https://github.com/facebookresearch/ResNeXt .
Transfer Learning was used in this task. Pre-trained weights of the model on the ImageNet dataset was used.
Along with transfer learning, data-augmentation, learning rate annealing, early stopping,etc. were also used in the training process.
This model was originally created for the kaggle challenge ( https://www.kaggle.com/c/qstp-deep-learning-2019 ) . The model acheived 3rd rank in the contest amongst 30 participants(Top 1 %).
The notebook contains the link to the weights of the final model, so that it can directly be used without training all over.
Image Classification Model in PyTorch for Indoor Scene Recognition
License
ashrutkumar/Indoor-scene-recognition
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Folders and files
Name | Name | Last commit message | Last commit date | |
---|---|---|---|---|
Repository files navigation
About
Image Classification Model in PyTorch for Indoor Scene Recognition
Topics
Resources
License
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published