fixing/refactoring cosy eval

docs/book/cosypose/evaluate.md

@@ -5,7 +5,7 @@ Please make sure you followed the steps relative to the evaluation in the main r
 Please run the following command to evaluate on YCBV dataset
 
 ```
-python -m happypose.pose_estimators.cosypose.cosypose.scripts.run_full_cosypose_eval_new detector_run_id=bop_pbr coarse_run_id=coarse-bop-ycbv-pbr--724183 refiner_run_id=refiner-bop-ycbv-pbr--604090 ds_names=["ycbv.bop19"] result_id=ycbv-debug detection_coarse_types=[["detector","S03_grid"]]
+python -m happypose.pose_estimators.cosypose.cosypose.scripts.run_full_cosypose_eval_new detector_run_id=bop_pbr coarse_run_id=coarse-bop-ycbv-pbr--724183 refiner_run_id=refiner-bop-ycbv-pbr--604090 ds_names=["ycbv.bop19"] result_id=ycbv-debug detection_coarse_types=[["detector"]]
 ```
 
 The other BOP datasets are supported as long as you download the correspond models.

happypose/pose_estimators/cosypose/cosypose/evaluation/evaluation.py

@@ -16,18 +16,11 @@
 from happypose.pose_estimators.cosypose.cosypose.evaluation.prediction_runner import (
     PredictionRunner,
 )
-from happypose.pose_estimators.cosypose.cosypose.integrated.detector import Detector
 from happypose.pose_estimators.cosypose.cosypose.integrated.pose_estimator import (
     PoseEstimator,
 )
 
 # Detection
-from happypose.pose_estimators.cosypose.cosypose.training.detector_models_cfg import (
-    check_update_config as check_update_config_detector,
-)
-from happypose.pose_estimators.cosypose.cosypose.training.detector_models_cfg import (
-    create_model_detector,
-)
 from happypose.pose_estimators.cosypose.cosypose.training.pose_models_cfg import (
     check_update_config as check_update_config_pose,
 )
@@ -45,8 +38,8 @@
 
 # Pose estimator
 from happypose.toolbox.datasets.datasets_cfg import make_object_dataset
+from happypose.toolbox.inference import load_detector
 from happypose.toolbox.lib3d.rigid_mesh_database import MeshDataBase
-from happypose.toolbox.renderer.panda3d_batch_renderer import Panda3dBatchRenderer
 from happypose.toolbox.utils.distributed import get_rank, get_tmp_dir
 from happypose.toolbox.utils.logging import get_logger
 
@@ -58,35 +51,50 @@
 logger = get_logger(__name__)
 
 
-def load_detector(run_id, ds_name):
-    run_dir = EXP_DIR / run_id
-    # cfg = yaml.load((run_dir / 'config.yaml').read_text(), Loader=yaml.FullLoader)
-    cfg = yaml.load((run_dir / "config.yaml").read_text(), Loader=yaml.UnsafeLoader)
-    cfg = check_update_config_detector(cfg)
-    label_to_category_id = cfg.label_to_category_id
-    model = create_model_detector(cfg, len(label_to_category_id))
-    ckpt = torch.load(run_dir / "checkpoint.pth.tar", map_location=device)
-    ckpt = ckpt["state_dict"]
-    model.load_state_dict(ckpt)
-    model = model.to(device).eval()
-    model.cfg = cfg
-    model.config = cfg
-    model = Detector(model, ds_name)
-    return model
-
-
-def load_pose_models(coarse_run_id, refiner_run_id, n_workers):
+# def load_detector(run_id, ds_name):
+#     run_dir = EXP_DIR / run_id
+#     # cfg = yaml.load((run_dir / 'config.yaml').read_text(), Loader=yaml.FullLoader)
+#     cfg = yaml.load((run_dir / "config.yaml").read_text(), Loader=yaml.UnsafeLoader)
+#     cfg = check_update_config_detector(cfg)
+#     label_to_category_id = cfg.label_to_category_id
+#     model = create_model_detector(cfg, len(label_to_category_id))
+#     ckpt = torch.load(run_dir / "checkpoint.pth.tar", map_location=device)
+#     ckpt = ckpt["state_dict"]
+#     model.load_state_dict(ckpt)
+#     model = model.to(device).eval()
+#     model.cfg = cfg
+#     model.config = cfg
+#     model = Detector(model, ds_name)
+#     return model
+
+
+def load_pose_models(coarse_run_id, refiner_run_id, n_workers, renderer_type="panda3d"):
     run_dir = EXP_DIR / coarse_run_id
-    # cfg = yaml.load((run_dir / 'config.yaml').read_text(), Loader=yaml.FullLoader)
     cfg = yaml.load((run_dir / "config.yaml").read_text(), Loader=yaml.UnsafeLoader)
     cfg = check_update_config_pose(cfg)
 
     object_dataset = make_object_dataset("ycbv")
-    renderer = Panda3dBatchRenderer(
-        object_dataset,
-        n_workers=n_workers,
-        preload_cache=False,
-    )
+    if renderer_type == "panda3d":
+        from happypose.toolbox.renderer.panda3d_batch_renderer import (
+            Panda3dBatchRenderer,
+        )
+
+        renderer = Panda3dBatchRenderer(
+            object_dataset,
+            n_workers=n_workers,
+            preload_cache=True,
+        )
+    elif renderer_type == "bullet":
+        from happypose.toolbox.renderer.bullet_batch_renderer import BulletBatchRenderer
+
+        renderer = BulletBatchRenderer(
+            object_dataset,
+            n_workers=n_workers,
+            preload_cache=True,
+        )
+    else:
+        raise ValueError(f"Renderer {renderer_type} not supported")
+
     mesh_db = MeshDataBase.from_object_ds(object_dataset)
     mesh_db_batched = mesh_db.batched().to(device)
 
@@ -173,7 +181,8 @@ def run_eval(
     # Load detector model
     if cfg.inference.detection_type == "detector":
         assert cfg.detector_run_id is not None
-        detector_model = load_detector(cfg.detector_run_id, cfg.ds_name)
+        # detector_model = load_detector(cfg.detector_run_id, cfg.ds_name)
+        detector_model = load_detector(cfg.detector_run_id)
     elif cfg.inference.detection_type == "gt":
         detector_model = None
     else:

happypose/pose_estimators/cosypose/cosypose/scripts/run_full_cosypose_eval_new.py

@@ -67,13 +67,11 @@
 def create_eval_cfg(
     cfg: EvalConfig,
     detection_type: str,
-    coarse_estimation_type: str,
     ds_name: str,
 ) -> Tuple[str, EvalConfig]:
     cfg = copy.deepcopy(cfg)
 
     cfg.inference.detection_type = detection_type
-    cfg.inference.coarse_estimation_type = coarse_estimation_type
     cfg.ds_name = ds_name
 
     if detection_type == "detector":
@@ -88,7 +86,7 @@ def create_eval_cfg(
         msg = f"Unknown detector type {cfg.detector_type}"
         raise ValueError(msg)
 
-    name = generate_save_key(detection_type, coarse_estimation_type)
+    name = generate_save_key(detection_type, "cosycoarse")
 
     return name, cfg
 
@@ -109,11 +107,10 @@ def run_full_eval(cfg: FullEvalConfig) -> None:
     for ds_name in cfg.ds_names:
         # create the EvalConfig objects that we will call `run_eval` on
         eval_configs: Dict[str, EvalConfig] = {}
-        for detection_type, coarse_estimation_type in cfg.detection_coarse_types:
+        for detection_type in cfg.detection_coarse_types:
             name, cfg_ = create_eval_cfg(
                 cfg,
                 detection_type,
-                coarse_estimation_type,
                 ds_name,
             )
             eval_configs[name] = cfg_

happypose/pose_estimators/megapose/inference/types.py

@@ -88,7 +88,7 @@ def assert_detections_valid(detections: DetectionsType) -> None:
 @dataclass
 class InferenceConfig:
     # TODO: move detection_type outside of here
-    detection_type: str = "detector"  # ['detector', 'gt']
+    detection_type: str = "detector"  # ['detector', 'gt', 'exte']
     coarse_estimation_type: str = "SO3_grid"
     SO3_grid_size: int = 576
     n_refiner_iterations: int = 5