Skip to content

Detects the class of food given in an image with pretrained mobilenetV2 and CNNs.

Notifications You must be signed in to change notification settings

kunal-bhadra/food-classification-CNN-mobilenetV2

Repository files navigation

Food Recognition with CNN

This is a project built while participating in the AIcrowd KIIT(AI) Mini Blitz Challenge and is the second one out of the three challenges given. We are given a train and test set and the objective is to detect the what food item is presen in the given image. This task only concerns with recognition and not localization, since we assume that there's one food class per image. For this, I went with CNNs and worked with about 9000+ datapoints.

🚀 The Result

I was able to score a F1 of 0.529 on the test set and secured the first place for this specific puzzle as of now.

✏ Tech Stack for Project Development

  • Python
  • Numpy
  • Pandas
  • Plotply
  • Seaborn
  • Scikit-learn
  • Tensorflow

🧠 The Approach

We used a Convolutional Neural Network with a pretrained model of Mobilenet_V2 and added two dense Keras layers on top which would, fine-tune it based on our specific dataset. For training, we trained for only 7 epochs and implemented early stopping to be efficient. At the end, the predictions on approximately 500 unlabelled images were submitted to get the above score.

🔗 Connect with me:

portfolio linkedin twitter

About

Detects the class of food given in an image with pretrained mobilenetV2 and CNNs.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published