Skip to content

Commit

Permalink
docs: update annotation
Browse files Browse the repository at this point in the history
Signed-off-by: Hayato Mizushima <[email protected]>
  • Loading branch information
hayato-m126 committed Sep 11, 2024
1 parent d09b91c commit c348ef5
Show file tree
Hide file tree
Showing 2 changed files with 12 additions and 4 deletions.
8 changes: 6 additions & 2 deletions docs/overview/index.en.md
Original file line number Diff line number Diff line change
Expand Up @@ -43,7 +43,11 @@ The details of the node's operation are shown in the figure below.
3. Create an evaluation scenario
1. Example scenarios could be found in the repository's [sample folder](https://github.com/tier4/driving_log_replayer_v2/tree/main/sample)
2. Refer to the [format definition](../result_format/index.md) section of this document for description contents.
4. If the node should test obstacle_segmentation, perception, perception_2d, or traffic_light stacks, please annotate with an annotation tool that supports conversion to t4_dataset.
4. Create a dataset
1. localization, eagleye, yabloc, ar_tag_based_localizer, and performance_diag are optional if you will not use [Evaluator](https://docs.web.auto/user-manuals/evaluator/introduction).
2. Create up to T4 non-annotated format data with reference to [perception_dataset tools_overview](https://github.com/tier4/tier4_perception_dataset/blob/main/docs/tools_overview.md).
3. If you create T4 non-annotated format data, it is possible to check the contents of the data set on [Vehicle Data Search](https://docs.web.auto/user-manuals/vehicle-data-search/quick-start#t4-dataset-%E3%81%AE%E5%8B%95%E7%94%BB%E8%A1%A8%E7%A4%BA).
5. If the node should test obstacle_segmentation, perception, perception_2d, or traffic_light stacks, please annotate with an annotation tool that supports conversion to t4_dataset.
1. [Deepen.AI](https://www.deepen.ai/) is available.
2. By adding conversion functionality to [perception_dataset](https://github.com/tier4/tier4_perception_dataset), it becomes possible to use other annotation tools as well.
5. Perform the evaluation.
6. Perform the evaluation.
8 changes: 6 additions & 2 deletions docs/overview/index.ja.md
Original file line number Diff line number Diff line change
Expand Up @@ -42,7 +42,11 @@ driving_log_replayer_v2 の評価ノードは、以下のように動作しま
3. シナリオを作成する
1. [sample folder](https://github.com/tier4/driving_log_replayer_v2/tree/main/sample) 内にシナリオの例あり
2. 記述内容は[フォーマット定義](../result_format/index.md)を参照
4. ユースケースが obstacle_segmentation, perception, perception_2d, traffic_light の場合、t4_dataset への変換に対応したアノテーションツールでアノテーションを実施する。
4. datasetを作成する
1. localization, eagleye, yabloc, ar_tag_based_localizer, performance_diagに関しては、[Evaluator](https://docs.web.auto/user-manuals/evaluator/introduction)を利用しないなら任意
2. [perception_dataset tools_overview](https://github.com/tier4/tier4_perception_dataset/blob/main/docs/tools_overview.md)を参考にT4 non-annotated format dataまで作る。
3. T4 non-annotated format dataまで作成すると、[Vehicle Data Search](https://docs.web.auto/user-manuals/vehicle-data-search/quick-start#t4-dataset-%E3%81%AE%E5%8B%95%E7%94%BB%E8%A1%A8%E7%A4%BA)上でデータセットの内容の確認が可能になる。
5. ユースケースが obstacle_segmentation, perception, perception_2d, traffic_light の場合、t4_dataset への変換に対応したアノテーションツールでアノテーションを実施する。
1. [Deepen.AI](https://www.deepen.ai/)が利用可能
2. [perception_dataset](https://github.com/tier4/tier4_perception_dataset)に変換機能を追加すれば他のアノテーションツールも使用可能になる
5. 評価を実行する。
6. 評価を実行する。

0 comments on commit c348ef5

Please sign in to comment.