Food Vector Art - Moozonian

Showing results for Food Vector Art

GitHub Repo https://github.com/cmcbride144/Food-Thesis

cmcbride144/Food-Thesis

In this thesis, we presented both behavioral and computational analyses of the linguocultural phenomenon of food. This was done by utilizing the skip-gram algorithm of state-of-the-art continuous space word representation models fastText (FaceBook) and Word2Vec (Google) and using the resulting vectors of food items as our independent variable in multi-linear regression. We found that we can predict the participant score of a food item based on the linguistic information embedded in the food item's vector with r = 0.65 and R2 = 0.73. This indicates that we can predict a person's culture, habits and traits based on the individual's textual representations.

GitHub Repo https://github.com/ElpinaFang/Food-Image-Classification-with-Swin-Transformer-and-SVM-Classifier

ElpinaFang/Food-Image-Classification-with-Swin-Transformer-and-SVM-Classifier

The dataset Food-101 is analyzed by evaluating and comparing the performance of five state-of-the-art models, they are MobileNetV3, ResNet50, EfficientNetV2, CoAtNet, and Swin Transformer. Subsequently, the research proceeds with the implementation of Support Vector Machine (SVM).