Avoid to keep in memory lengths and bins for ImplicitronRayBundle

Summary:
Convert ImplicitronRayBundle to a "classic" class instead of a dataclass. This change is introduced as a way to preserve the ImplicitronRayBundle interface while allowing two outcomes:
- init lengths arguments is now a Optional[torch.Tensor] instead of torch.Tensor
- lengths is now a property which returns a `torch.Tensor`. The lengths property will either recompute lengths from bins or return the stored _lengths. `_lenghts` is None if bins is set. It saves us a bit of memory.

Reviewed By: shapovalov

Differential Revision: D46686094

fbshipit-source-id: 3c75c0947216476ebff542b6f552d311024a679b

pytorch3d/implicitron/models/renderer/base.py

@@ -6,8 +6,6 @@
 
 from __future__ import annotations
 
-import dataclasses
-
 from abc import ABC, abstractmethod
 from dataclasses import dataclass, field
 from enum import Enum
@@ -29,7 +27,6 @@ class RenderSamplingMode(Enum):
     FULL_GRID = "full_grid"
 
 
-@dataclasses.dataclass
 class ImplicitronRayBundle:
     """
     Parametrizes points along projection rays by storing ray `origins`,
@@ -69,53 +66,58 @@ class ImplicitronRayBundle:
             lengths should be equal to the midpoints of bins `(..., num_points_per_ray)`.
         pixel_radii_2d: An optional tensor of shape `(..., 1)`
             base radii of the conical frustums.
+
+    Raises:
+        ValueError: If either bins or lengths are not provided.
+        ValueError: If bins is provided and the last dim is inferior or equal to 1.
     """
 
-    origins: torch.Tensor
-    directions: torch.Tensor
-    lengths: torch.Tensor
-    xys: torch.Tensor
-    camera_ids: Optional[torch.LongTensor] = None
-    camera_counts: Optional[torch.LongTensor] = None
-    bins: Optional[torch.Tensor] = None
-    pixel_radii_2d: Optional[torch.Tensor] = None
-
-    @classmethod
-    def from_bins(
-        cls,
+    def __init__(
+        self,
         origins: torch.Tensor,
         directions: torch.Tensor,
-        bins: torch.Tensor,
+        lengths: Optional[torch.Tensor],
         xys: torch.Tensor,
-        **kwargs,
-    ) -> "ImplicitronRayBundle":
-        """
-        Creates a new instance from bins instead of lengths.
-
-        Attributes:
-            origins: A tensor of shape `(..., 3)` denoting the
-                origins of the sampling rays in world coords.
-            directions: A tensor of shape `(..., 3)` containing the direction
-                vectors of sampling rays in world coords. They don't have to be normalized;
-                they define unit vectors in the respective 1D coordinate systems; see
-                documentation for :func:`ray_bundle_to_ray_points` for the conversion formula.
-            bins: A tensor of shape `(..., num_points_per_ray + 1)`
-                containing the bins at which the rays are sampled. In this case
-                lengths is equal to the midpoints of bins `(..., num_points_per_ray)`.
-            xys: A tensor of shape `(..., 2)`, the xy-locations (`xys`) of the ray pixels
-            kwargs: Additional arguments passed to the constructor of ImplicitronRayBundle
-        Returns:
-            An instance of ImplicitronRayBundle.
-        """
-
-        if bins.shape[-1] <= 1:
+        camera_ids: Optional[torch.LongTensor] = None,
+        camera_counts: Optional[torch.LongTensor] = None,
+        bins: Optional[torch.Tensor] = None,
+        pixel_radii_2d: Optional[torch.Tensor] = None,
+    ):
+        if bins is not None and bins.shape[-1] <= 1:
             raise ValueError(
                 "The last dim of bins must be at least superior or equal to 2."
             )
-        # equivalent to: 0.5 * (bins[..., 1:] + bins[..., :-1]) but more efficient
-        lengths = torch.lerp(bins[..., 1:], bins[..., :-1], 0.5)
 
-        return cls(origins, directions, lengths, xys, bins=bins, **kwargs)
+        if bins is None and lengths is None:
+            raise ValueError(
+                "Please set either bins or lengths to initialize an ImplicitronRayBundle."
+            )
+
+        self.origins = origins
+        self.directions = directions
+        self._lengths = lengths if bins is None else None
+        self.xys = xys
+        self.bins = bins
+        self.pixel_radii_2d = pixel_radii_2d
+        self.camera_ids = camera_ids
+        self.camera_counts = camera_counts
+
+    @property
+    def lengths(self) -> torch.Tensor:
+        if self.bins is not None:
+            # equivalent to: 0.5 * (bins[..., 1:] + bins[..., :-1]) but more efficient
+            # pyre-ignore
+            return torch.lerp(self.bins[..., :-1], self.bins[..., 1:], 0.5)
+        return self._lengths
+
+    @lengths.setter
+    def lengths(self, value):
+        if self.bins is not None:
+            raise ValueError(
+                "If the bins attribute is not None you cannot set the lengths attribute."
+            )
+        else:
+            self._lengths = value
 
     def is_packed(self) -> bool:
         """

pytorch3d/implicitron/models/renderer/lstm_renderer.py

@@ -4,7 +4,7 @@
 # This source code is licensed under the BSD-style license found in the
 # LICENSE file in the root directory of this source tree.
 
-import dataclasses
+import copy
 import logging
 from typing import List, Optional, Tuple
 
@@ -102,12 +102,11 @@ def forward(
             )
 
         # jitter the initial depths
-        ray_bundle_t = dataclasses.replace(
-            ray_bundle,
-            lengths=(
-                ray_bundle.lengths
-                + torch.randn_like(ray_bundle.lengths) * self.init_depth_noise_std
-            ),
+
+        ray_bundle_t = copy.copy(ray_bundle)
+        ray_bundle_t.lengths = (
+            ray_bundle.lengths
+            + torch.randn_like(ray_bundle.lengths) * self.init_depth_noise_std
         )
 
         states: List[Optional[Tuple[torch.Tensor, torch.Tensor]]] = [None]

pytorch3d/implicitron/models/renderer/ray_point_refiner.py

@@ -4,6 +4,8 @@
 # This source code is licensed under the BSD-style license found in the
 # LICENSE file in the root directory of this source tree.
 
+import copy
+
 import torch
 from pytorch3d.implicitron.models.renderer.base import ImplicitronRayBundle
 from pytorch3d.implicitron.tools.config import Configurable, expand_args_fields
@@ -106,14 +108,13 @@ def forward(
             z_vals = z_samples
         # Resort by depth.
         z_vals, _ = torch.sort(z_vals, dim=-1)
-
-        kwargs_ray = dict(vars(input_ray_bundle))
+        ray_bundle = copy.copy(input_ray_bundle)
         if input_ray_bundle.bins is None:
-            kwargs_ray["lengths"] = z_vals
-            return ImplicitronRayBundle(**kwargs_ray)
-        kwargs_ray["bins"] = z_vals
-        del kwargs_ray["lengths"]
-        return ImplicitronRayBundle.from_bins(**kwargs_ray)
+            ray_bundle.lengths = z_vals
+        else:
+            ray_bundle.bins = z_vals
+
+        return ray_bundle
 
 
 def apply_blurpool_on_weights(weights) -> torch.Tensor:

pytorch3d/implicitron/models/renderer/ray_sampler.py

@@ -236,11 +236,12 @@ def forward(
         elif self.cast_ray_bundle_as_cone:
             pixel_hw: Tuple[float, float] = (self.pixel_height, self.pixel_width)
             pixel_radii_2d = compute_radii(cameras, ray_bundle.xys[..., :2], pixel_hw)
-            return ImplicitronRayBundle.from_bins(
+            return ImplicitronRayBundle(
                 directions=ray_bundle.directions,
                 origins=ray_bundle.origins,
-                bins=ray_bundle.lengths,
+                lengths=None,
                 xys=ray_bundle.xys,
+                bins=ray_bundle.lengths,
                 pixel_radii_2d=pixel_radii_2d,
             )
 

tests/implicitron/test_models_renderer_base.py

@@ -25,23 +25,62 @@
 class TestRendererBase(TestCaseMixin, unittest.TestCase):
     def test_implicitron_from_bins(self) -> None:
         bins = torch.randn(2, 3, 4, 5)
-        ray_bundle = ImplicitronRayBundle.from_bins(
+        ray_bundle = ImplicitronRayBundle(
             origins=None,
             directions=None,
+            lengths=None,
             xys=None,
             bins=bins,
         )
         self.assertClose(ray_bundle.lengths, 0.5 * (bins[..., 1:] + bins[..., :-1]))
         self.assertClose(ray_bundle.bins, bins)
 
+    def test_implicitron_raise_value_error_bins_is_set_and_try_to_set_lengths(
+        self,
+    ) -> None:
+        with self.assertRaises(ValueError) as context:
+            ray_bundle = ImplicitronRayBundle(
+                origins=torch.rand(2, 3, 4, 3),
+                directions=torch.rand(2, 3, 4, 3),
+                lengths=None,
+                xys=torch.rand(2, 3, 4, 2),
+                bins=torch.rand(2, 3, 4, 1),
+            )
+            ray_bundle.lengths = torch.empty(2)
+            self.assertEqual(
+                str(context.exception),
+                "If the bins attribute is not None you cannot set the lengths attribute.",
+            )
+
     def test_implicitron_raise_value_error_if_bins_dim_equal_1(self) -> None:
-        with self.assertRaises(ValueError):
-            ImplicitronRayBundle.from_bins(
+        with self.assertRaises(ValueError) as context:
+            ImplicitronRayBundle(
                 origins=torch.rand(2, 3, 4, 3),
                 directions=torch.rand(2, 3, 4, 3),
+                lengths=None,
                 xys=torch.rand(2, 3, 4, 2),
                 bins=torch.rand(2, 3, 4, 1),
             )
+            self.assertEqual(
+                str(context.exception),
+                "The last dim of bins must be at least superior or equal to 2.",
+            )
+
+    def test_implicitron_raise_value_error_if_neither_bins_or_lengths_provided(
+        self,
+    ) -> None:
+        with self.assertRaises(ValueError) as context:
+            ImplicitronRayBundle(
+                origins=torch.rand(2, 3, 4, 3),
+                directions=torch.rand(2, 3, 4, 3),
+                lengths=None,
+                xys=torch.rand(2, 3, 4, 2),
+                bins=None,
+            )
+            self.assertEqual(
+                str(context.exception),
+                "Please set either bins or lengths to initialize an ImplicitronRayBundle.",
+            )
 
     def test_conical_frustum_to_gaussian(self) -> None:
         origins = torch.zeros(3, 3, 3)