wrapper for uploading LSS predictions (#436)

Author: jean-lucas

.circleci/config.yml

@@ -60,7 +60,7 @@ jobs:
             mkdir test_results
             set -e
             TEST_FILES=$(circleci tests glob "tests/**/test_*.py")
-            echo "$TEST_FILES" | circleci tests run --command "xargs poetry run coverage run --include=nucleus/* -m pytest -n auto -s -v -o junit_family=legacy --junitxml=test_results/junit.xml" --verbose --split-by=timings
+            echo "$TEST_FILES" | circleci tests run --command "xargs poetry run coverage run --include=nucleus/* -m pytest -s -v -o junit_family=legacy --junitxml=test_results/junit.xml" --verbose --split-by=timings
             poetry run coverage report
             poetry run coverage html
       - store_test_results:

CHANGELOG.md

@@ -5,6 +5,25 @@ All notable changes to the [Nucleus Python Client](https://github.com/scaleapi/n
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [0.17.5](https://github.com/scaleapi/nucleus-python-client/releases/tag/v0.17.5) - 2024-04-15
+
+### Added
+- Method for uploading lidar semantic segmentation predictions, via `dataset.upload_lidar_semseg_predictions`
+
+Example usage:
+
+```python
+dataset = client.get_dataset("ds_...")
+model = client.get_model("prj_...")
+pointcloud_ref_id = 'pc_ref_1'
+predictions_s3 = "s3://temp/predictions.json"
+
+dataset.upload_lidar_semseg_predictions(model, pointcloud_ref_id, predictions_s3)
+```
+
+For the expected format of the s3 predictions, refer to the [documentation here](https://docs.nucleus.scale.com/en/latest/api/nucleus/index.html#nucleus.Dataset.upload_lidar_semseg_predictions)
+
+
 ## [0.17.4](https://github.com/scaleapi/nucleus-python-client/releases/tag/v0.17.4) - 2024-03-25
 
 ### Modified

nucleus/dataset.py

@@ -104,7 +104,7 @@
 from .upload_response import UploadResponse
 
 if TYPE_CHECKING:
-    from . import NucleusClient
+    from . import Model, NucleusClient
 
 # TODO: refactor to reduce this file to under 1000 lines.
 # pylint: disable=C0302
@@ -2303,3 +2303,41 @@ def add_items_from_dir(
 
         else:
             print(f"Did not find any items in {dirname}.")
+
+    def upload_lidar_semseg_predictions(
+        self,
+        model: "Model",
+        pointcloud_ref_id: str,
+        predictions_s3_path: str,
+    ):
+        """Upload Lidar Semantic Segmentation predictions for a given pointcloud.
+
+        Assuming a pointcloud with only 4 points (three labeled as Car, one labeled as Person),
+        the contents of the predictions s3 object should be formatted as such:
+
+        .. code-block:: json
+
+            {
+                "objects": [
+                    { "label": "Car", "index": 1},
+                    { "label": "Person", "index": 2}
+                ],
+                "point_objects": [1, 1, 1, 2],
+                "point_confidence": [0.5, 0.9, 0.9, 0.3]
+            }
+
+        The order of the points in the `"point_objects"` should be in the same order as the points that
+        were originally uploaded to Scale.
+
+        Parameters:
+            model (:class:`Model`): Nucleus model used to store these predictions
+            pointcloud_ref_id (str): The reference ID of the pointcloud for which these predictions belong to
+            predictions_s3_path (str): S3 path to where the predictions are stored
+
+        """
+
+        return self._client.make_request(
+            payload={"pointsSegmentationUrl": predictions_s3_path},
+            route=f"dataset/{self.id}/model/{model.id}/pointcloud/{pointcloud_ref_id}/uploadLSSPrediction",
+            requests_command=requests.post,
+        )

pyproject.toml

@@ -25,7 +25,7 @@ ignore = ["E501", "E741", "E731", "F401"]  # Easy ignore for getting it running
 
 [tool.poetry]
 name = "scale-nucleus"
-version = "0.17.4"
+version = "0.17.5"
 description = "The official Python client library for Nucleus, the Data Platform for AI"
 license =  "MIT"
 authors = ["Scale AI Nucleus Team <[email protected]>"]