Kitchenware

The goal of the competition was to classify images of kitchenware items into 6 classes: - cups - glasses - plates - spoons - forks - knives

Experiments with Effiecentnet using TensorFlow and Keras
- learning rates
- l2 regularization
- using different versions of the Effiecentnet
Utilized synthetic data to even out class imbalance
Incorporated imgaug
- to generate synthetic data
- improve training by presenting the neural network with variations of the training images

Name		Name	Last commit message	Last commit date
Latest commit History 40 Commits
.gitignore		.gitignore
LICENSE		LICENSE
README.md		README.md
cups.ipynb		cups.ipynb
eda.ipynb		eda.ipynb
effb2_imnet.ipynb		effb2_imnet.ipynb
effb2_imnet_aug_v2.ipynb		effb2_imnet_aug_v2.ipynb
effb2_imnet_v2.ipynb		effb2_imnet_v2.ipynb
effb2_imnet_v3.ipynb		effb2_imnet_v3.ipynb
effb2_imnet_v4.ipynb		effb2_imnet_v4.ipynb
effb2_imnet_v4_aug.ipynb		effb2_imnet_v4_aug.ipynb
effb2_imnet_v4_more_forks.ipynb		effb2_imnet_v4_more_forks.ipynb
effb2_imnet_v5.ipynb		effb2_imnet_v5.ipynb
effb2_imnet_v8_more_forks.ipynb		effb2_imnet_v8_more_forks.ipynb
effb2_imnet_v8_more_forks_glasses.ipynb		effb2_imnet_v8_more_forks_glasses.ipynb
effb2_imnet_v8_more_forks_glasses_b7.ipynb		effb2_imnet_v8_more_forks_glasses_b7.ipynb
effb2_imnet_v9_more_forks_glasses_b4.ipynb		effb2_imnet_v9_more_forks_glasses_b4.ipynb
effbv2l_imnet_v9_more_forks_glasses.ipynb		effbv2l_imnet_v9_more_forks_glasses.ipynb
effbv2l_imnet_v9_more_forks_glasses_aug.ipynb		effbv2l_imnet_v9_more_forks_glasses_aug.ipynb
forks.ipynb		forks.ipynb
glasses.ipynb		glasses.ipynb
knives.ipynb		knives.ipynb
plates.ipynb		plates.ipynb
sample_submission.ipynb		sample_submission.ipynb
sample_submission_extra_images.ipynb		sample_submission_extra_images.ipynb
spoons.ipynb		spoons.ipynb

Provide feedback