Explored the concept of UDA by training a model (on the MNIST dataset) with and without UDA and evaluated both of them on the Coloured MNIST dataset.
Found that model with UDA performed much better than the vanilla classification model
| Name | Name | Last commit date | ||
|---|---|---|---|---|