Antonin Vidon's Projects
Website.
Classification project aiming at predicting whether a breast tissue patch (scanned at x40) is cancerous.
I went back to my previous project and experimented with custom ResNet architectures. The classification task aims at predicting whether a breast tissue patch (scanned at x40) is cancerous.
We carried out thematic clustering and differential prediction of number of retweets with Gradient Boosting and Quantile regression. We also performed text embedding with Bidirectional Encoder Representations (BERT, Google) for deep prediction with TensorFlow
Using visualization to put to the test some intuitive insights between energy consumption and human development.
An R package meant to create animated gifs from a dynamic interface.
This repo contains our attempt to replicate the results of pix2pix on the 'Facades' dataset as well on a dataset we created from Country211 data.
Robust Generalization in Automatic Speech Recognition: wav2vec 2.0 and Whisper
Implementation of SE-ResNet, SE-ResNeXt and SE-InceptionV3 from scratch and comparison of the results obtained for CIFAR-10, CIFAR-100 and Tiny ImageNet with the original paper.
Developed phase recognition models based on MobileNetV2 to classify frames from Hernia surgery videos.