nicolas-dufour

plonk. Code for "Around the World in 80 Timesteps: A Generative Approach to Global Visual Geolocation"

CAD. Code for "Don’t drop your samples! Coherence-aware training benefits Conditional diffusion" CVPR 2024 Highlight

SCAM. Code for SCAM! Transferring humans between images with Semantic Cross Attention Modulation. Also contains implementation for SPADE, CLADE, SEAN and INADE

miro. Code for MIRO: MultI-Reward cOnditioned pretraining improves T2I quality and efficiency

time-series-self-supervised-learning. This project implements a self supervised learning method for time series

modded-nanogpt-SOAP-pom. NanoGPT (124M) quality in 3.7B tokens with SOAP optimizer

self-supervised-low-res-speech. This project transfert the self supervised Wav2vec2 representation to low ressources languages

instance-video-colorization. This project aims to use the instance colorization model and apply it to video in a time consistent manner

nicolas-dufour.github.io. JavaScript

bird-classification-challenge. This is my solution for the RecVis Challenge where i ended 1st on fine-grained classification on bird dataset

slides_phd_defense. JavaScript

datascientist-job-study. Study done in 2019 on the data scientist job

cheese_classification_challenge. Baseline code for INF473V challenge for cheese classification

dnn-from-scratch. Implemented Neural networks, RBM and DBN from scratch