Week 2 · Day 9 · Thursday 22 October

Image Classification and Recognition

Recognition beyond a fixed label set: adapting pretrained classifiers, fine-grained recognition, metric learning, retrieval and re-identification, and rejecting images of classes the model has never seen.

Topic Description & Objectives Resource
Image Classification and Recognition Adapting pretrained classifiers, fine-grained recognition, deep metric learning, image retrieval, re-identification, and open-set recognition and out-of-distribution detection Open
Lab 1: Out-of-Distribution Detection Train a CIFAR-10 classifier, then compare softmax confidence and energy scores at flagging CIFAR-100, SVHN and texture images, measured with AUROC and FPR95 Notebook Colab