suppress errors in vision/fair/detectron2/projects/DensePose

Differential Revision: D33202211

fbshipit-source-id: 10fb961968bd944783ec229941d412bbe7e4afa8

Author: Pyre Bot Jr

projects/DensePose/densepose/converters/chart_output_to_chart_result.py

@@ -36,7 +36,9 @@ def resample_uv_tensors_to_bbox(
     x, y, w, h = box_xywh_abs
     w = max(int(w), 1)
     h = max(int(h), 1)
+    # pyre-fixme[6]: Expected `Optional[int]` for 2nd param but got `Tuple[int, int]`.
     u_bbox = F.interpolate(u, (h, w), mode="bilinear", align_corners=False)
+    # pyre-fixme[6]: Expected `Optional[int]` for 2nd param but got `Tuple[int, int]`.
     v_bbox = F.interpolate(v, (h, w), mode="bilinear", align_corners=False)
     uv = torch.zeros([2, h, w], dtype=torch.float32, device=u.device)
     for part_id in range(1, u_bbox.size(1)):
@@ -137,6 +139,8 @@ def resample_confidences_to_bbox(
     # assign data from channels that correspond to the labels
     for key in confidence_names:
         resampled_confidence = F.interpolate(
+            # pyre-fixme[6]: Expected `Optional[int]` for 2nd param but got
+            #  `Tuple[int, int]`.
             getattr(predictor_output, key), (h, w), mode="bilinear", align_corners=False
         )
         result = confidence_base.clone()

projects/DensePose/densepose/converters/segm_to_mask.py

@@ -25,6 +25,8 @@ def resample_coarse_segm_tensor_to_bbox(coarse_segm: torch.Tensor, box_xywh_abs:
     x, y, w, h = box_xywh_abs
     w = max(int(w), 1)
     h = max(int(h), 1)
+    # pyre-fixme[16]: `Tensor` has no attribute `argmax`.
+    # pyre-fixme[6]: Expected `Optional[int]` for 2nd param but got `Tuple[int, int]`.
     labels = F.interpolate(coarse_segm, (h, w), mode="bilinear", align_corners=False).argmax(dim=1)
     return labels
 
@@ -48,11 +50,16 @@ def resample_fine_and_coarse_segm_tensors_to_bbox(
     w = max(int(w), 1)
     h = max(int(h), 1)
     # coarse segmentation
+    # pyre-fixme[16]: `Tensor` has no attribute `argmax`.
     coarse_segm_bbox = F.interpolate(
+        # pyre-fixme[6]: Expected `Optional[int]` for 2nd param but got `Tuple[int,
+        #  int]`.
         coarse_segm, (h, w), mode="bilinear", align_corners=False
     ).argmax(dim=1)
     # combined coarse and fine segmentation
     labels = (
+        # pyre-fixme[6]: Expected `Optional[int]` for 2nd param but got `Tuple[int,
+        #  int]`.
         F.interpolate(fine_segm, (h, w), mode="bilinear", align_corners=False).argmax(dim=1)
         * (coarse_segm_bbox > 0).long()
     )

projects/DensePose/densepose/data/samplers/densepose_base.py

@@ -141,13 +141,18 @@ def _resample_mask(self, output: Any) -> torch.Tensor:
         """
         sz = DensePoseDataRelative.MASK_SIZE
         S = (
+            # pyre-fixme[16]: `Tensor` has no attribute `argmax`.
+            # pyre-fixme[6]: Expected `Optional[int]` for 2nd param but got
+            #  `Tuple[int, int]`.
             F.interpolate(output.coarse_segm, (sz, sz), mode="bilinear", align_corners=False)
             .argmax(dim=1)
             .long()
         )
         I = (
             (
                 F.interpolate(
+                    # pyre-fixme[6]: Expected `Optional[int]` for 2nd param but got
+                    #  `Tuple[int, int]`.
                     output.fine_segm, (sz, sz), mode="bilinear", align_corners=False
                 ).argmax(dim=1)
                 * (S > 0).long()

projects/DensePose/densepose/data/samplers/densepose_cse_base.py

@@ -109,7 +109,11 @@ def _produce_mask_and_results(
         S = densepose_output.coarse_segm
         E = densepose_output.embedding
         _, _, w, h = bbox_xywh
+        # pyre-fixme[6]: Expected `Optional[int]` for 2nd param but got `Tuple[int,
+        #  int]`.
         embeddings = F.interpolate(E, size=(h, w), mode="bilinear")[0]
+        # pyre-fixme[6]: Expected `Optional[int]` for 2nd param but got `Tuple[int,
+        #  int]`.
         coarse_segm_resized = F.interpolate(S, size=(h, w), mode="bilinear")[0]
         mask = coarse_segm_resized.argmax(0) > 0
         other_values = torch.empty((0, h, w), device=E.device)
@@ -130,6 +134,9 @@ def _resample_mask(self, output: Any) -> torch.Tensor:
         """
         sz = DensePoseDataRelative.MASK_SIZE
         mask = (
+            # pyre-fixme[16]: `Tensor` has no attribute `argmax`.
+            # pyre-fixme[6]: Expected `Optional[int]` for 2nd param but got
+            #  `Tuple[int, int]`.
             F.interpolate(output.coarse_segm, (sz, sz), mode="bilinear", align_corners=False)
             .argmax(dim=1)
             .long()

projects/DensePose/densepose/data/samplers/densepose_cse_confidence_based.py

@@ -112,6 +112,8 @@ def _produce_mask_and_results(
         densepose_output = instance.pred_densepose
         mask, embeddings, _ = super()._produce_mask_and_results(instance, bbox_xywh)
         other_values = F.interpolate(
+            # pyre-fixme[6]: Expected `Optional[int]` for 2nd param but got
+            #  `Tuple[int, int]`.
             getattr(densepose_output, self.confidence_channel), size=(h, w), mode="bilinear"
         )[0].cpu()
         return mask, embeddings, other_values

projects/DensePose/densepose/data/transform/image.py

@@ -31,6 +31,8 @@ def __call__(self, images: torch.Tensor) -> torch.Tensor:
         max_size = max(images.shape[-2:])
         scale = min(self.min_size / min_size, self.max_size / max_size)
         images = torch.nn.functional.interpolate(
+            # pyre-fixme[6]: Expected `Optional[typing.List[float]]` for 2nd param
+            #  but got `float`.
             images, scale_factor=scale, mode="bilinear", align_corners=False
         )
         return images

projects/DensePose/densepose/engine/trainer.py

@@ -77,7 +77,6 @@ def extract_embedder_from_model(cls, model: nn.Module) -> Optional[Embedder]:
         if isinstance(model, nn.parallel.DistributedDataParallel):
             model = model.module
         if hasattr(model, "roi_heads") and hasattr(model.roi_heads, "embedder"):
-            # pyre-fixme[16]: `Tensor` has no attribute `embedder`.
             return model.roi_heads.embedder
         return None
 

projects/DensePose/densepose/evaluation/densepose_coco_evaluation.py

@@ -543,7 +543,10 @@ def _extract_mask(self, dt: Dict[str, Any]) -> np.ndarray:
             dy = max(int(dt["bbox"][3]), 1)
             dx = max(int(dt["bbox"][2]), 1)
             return (
+                # pyre-fixme[16]: `Tensor` has no attribute `argmax`.
                 F.interpolate(
+                    # pyre-fixme[6]: Expected `Optional[int]` for 2nd param but got
+                    #  `Tuple[int, int]`.
                     dt["coarse_segm"].unsqueeze(0), (dy, dx), mode="bilinear", align_corners=False
                 )
                 .squeeze(0)
@@ -561,6 +564,8 @@ def _extract_mask(self, dt: Dict[str, Any]) -> np.ndarray:
             dx = max(int(dt["bbox"][2]), 1)
             return (
                 F.interpolate(
+                    # pyre-fixme[6]: Expected `Optional[int]` for 2nd param but got
+                    #  `Tuple[int, int]`.
                     coarse_segm.unsqueeze(0), (dy, dx), mode="bilinear", align_corners=False
                 )
                 .squeeze(0)

projects/DensePose/densepose/modeling/cse/embedder.py

@@ -58,7 +58,6 @@ def create_embedder(embedder_spec: CfgNode, embedder_dim: int) -> nn.Module:
         raise ValueError(f"Unexpected embedder type {embedder_type}")
 
     if not embedder_spec.IS_TRAINABLE:
-        # pyre-fixme[29]: `Union[nn.Module, torch.Tensor]` is not a function.
         embedder.requires_grad_(False)
 
     return embedder
@@ -86,7 +85,6 @@ def __init__(self, cfg: CfgNode):
         logger = logging.getLogger(__name__)
         for mesh_name, embedder_spec in cfg.MODEL.ROI_DENSEPOSE_HEAD.CSE.EMBEDDERS.items():
             logger.info(f"Adding embedder embedder_{mesh_name} with spec {embedder_spec}")
-            # pyre-fixme[29]: `Union[nn.Module, torch.Tensor]` is not a function.
             self.add_module(f"embedder_{mesh_name}", create_embedder(embedder_spec, embedder_dim))
             self.mesh_names.add(mesh_name)
         if cfg.MODEL.WEIGHTS != "":

projects/DensePose/densepose/modeling/cse/utils.py

@@ -63,7 +63,9 @@ def get_closest_vertices_mask_from_ES(
         Closest Vertices (tensor [h, w]), int, for every point of the resulting box
         Segmentation mask (tensor [h, w]), boolean, for every point of the resulting box
     """
+    # pyre-fixme[6]: Expected `Optional[int]` for 2nd param but got `Tuple[int, int]`.
     embedding_resized = F.interpolate(E, size=(h, w), mode="bilinear")[0].to(device)
+    # pyre-fixme[6]: Expected `Optional[int]` for 2nd param but got `Tuple[int, int]`.
     coarse_segm_resized = F.interpolate(S, size=(h, w), mode="bilinear")[0].to(device)
     mask = coarse_segm_resized.argmax(0) > 0
     closest_vertices = torch.zeros(mask.shape, dtype=torch.long, device=device)