This section is included if you are curious about what has changed between MMSeg 0.x and 1.x.
<table>
<tr>
<td>MMSegmentation 0.x</td>
<td>MMSegmentation 1.x</td>
</tr>
<tr>
<td>mmseg.api</td>
<td>mmseg.api</td>
</tr>
<tr>
<tdbgcolor=#fcf7f7>- mmseg.core</td>
<tdbgcolor=#ecf4eb>+ mmseg.engine</td>
</tr>
<tr>
<td>mmseg.datasets</td>
<td>mmseg.datasets</td>
</tr>
<tr>
<td>mmseg.models</td>
<td>mmseg.models</td>
</tr>
<tr>
<tdbgcolor=#fcf7f7>- mmseg.ops</td>
<tdbgcolor=#ecf4eb>+ mmseg.structure</td>
</tr>
<tr>
<td>mmseg.utils</td>
<td>mmseg.utils</td>
</tr>
<tr>
<td></td>
<tdbgcolor=#ecf4eb>+ mmseg.evaluation</td>
</tr>
<tr>
<td></td>
<tdbgcolor=#ecf4eb>+ mmseg.registry</td>
<tr>
</table>
## Removed packages
### `mmseg.core`
In OpenMMLab 2.0, `core` package has been removed. `hooks` and `optimizers` of `core` are moved in `mmseg.engine`, and `evaluation` in `core` is mmseg.evaluation currently.
## `mmseg.ops`
`ops` package included `encoding` and `wrappers`, which are moved in `mmseg.models.utils`.
## Added packages
### `mmseg.engine`
OpenMMLab 2.0 adds a new foundational library for training deep learning, MMEngine. It servers as the training engine of all OpenMMLab codebases.
`engine` package of mmseg is some customized modules for semantic segmentation task, like `SegVisualizationHook` which works for visualizing segmentation mask.
### `mmseg.structure`
In OpenMMLab 2.0, we designed data structure for computer vision task, and in mmseg, we implements `SegDataSample` in `structure` package.
### `mmseg.evaluation`
We move all evaluation metric in `mmseg.evaluation`.
### `mmseg.registry`
We moved registry implementations for all kinds of modules in MMSegmentation in `mmseg.registry`.
OpenMMLab 2.0 tries to support unified interface for multitasking of Computer Vision, and releases much stronger [`Runner`](https://github.com/open-mmlab/mmengine/blob/main/docs/en/design/runner.md), so MMSeg 1.x removed modules in `train.py` and `test.py` renamed `init_segmentor` to `init_model` and `inference_segmentor` to `inference_model`.
OpenMMLab 2.0 defines the `BaseDataset` to function and interface of dataset, and MMSegmentation 1.x also follow this protocol and defines the `BaseSegDataset` inherited from `BaseDataset`. MMCV 2.x collects general data transforms for multiple tasks e.g. classification, detection, segmentation, so MMSegmentation 1.x uses these data transforms and removes them from mmseg.datasets.
