Deep Learning Systems Project Files
-
each dataset image in different directory with naming convention: dataset_{id}_images
-
each dataset csv a different csv with columns: = ['Title','Instructions','Image_Name','Cleaned_Ingredients','dataset','id']
-
each csv name: cleaned_dataset_{id}.csv
-
dataset URL: https://www.kaggle.com/alphadraco/deepfoodie
- Title: lower cased, no special characters
- Instructions: lower cased, no word numbers
- Cleaned_Ingredients: only words no numbers, probably from a many to one dictionary
- Resnet50, EfficientNet
- Different embeddings: Glove, embeddings from food datasets, 1M+ embeddings
- Prediction Ranker