b-ptiste/sketches-classification

sketches-classification

This project was carried out as part of the Reconnaissance d'objets et vision artificielle (RecVis) - Master M2 MVA Lecturers: Gül Varol, Jean Ponce, Armand Joulin, Josef Sivic, Ivan Laptev, Cordelia Schmid, and Mathieu Aubry

Kaggle (rank 4/59) : https://www.kaggle.com/competitions/mva-recvis-2023/leaderboard

The aim of this work is to develop a model capable of classifying the images of the dataset classifysketch with the best accuracy. It is made up of 250 classes of sketches. We will begin by examining the dataset, then discuss the model selection, data augmentation and model tuning that enabled me to achieve 82.8% accuracy on the test dataset using results from [1] and [2] and a new data augmentation

• Implementation of Cut-mix, Mix-up and Line width augmentation :

Credit

• Teacher assistant : Ricardo Garcia Pinel