Add Ray Tracing (pytorch#3604) (pytorch#2850) (pytorch#3655)

Summary:
Revamped version of pytorch#3234
(which was also revamp of pytorch#2850)

docs/source/prototype.functional.rst

@@ -35,4 +35,5 @@ Room Impulse Response Simulation
    :toctree: generated
    :nosignatures:
 
+   ray_tracing
    simulate_rir_ism

src/libtorchaudio/rir/ray_tracing.cpp

@@ -220,17 +220,21 @@ class RayTracer {
           if (NORM(to_mic - dir * impact_distance) < mic_radius + EPS) {
             // The length of this last hop
             auto travel_dist_at_mic = travel_dist + std::abs(impact_distance);
+            auto bin_idx = get_bin_idx(travel_dist_at_mic);
+            if (bin_idx >= histograms.size(1)) {
+              continue;
+            }
             auto coeff = get_energy_coeff(travel_dist_at_mic, mic_radius_sq);
             auto energy = energies / coeff;
-            histograms[mic_idx][get_bin_idx(travel_dist_at_mic)] += energy;
+            histograms[mic_idx][bin_idx] += energy;
           }
         }
       }
 
       travel_dist += hit_distance;
       energies *= wall.reflection;
 
-      //   Let's shoot the scattered ray induced by the rebound on the wall
+      // Let's shoot the scattered ray induced by the rebound on the wall
       if (do_scattering) {
         scat_ray(histograms, wall, energies, origin, hit_point, travel_dist);
         energies *= (1. - wall.scattering);

src/libtorchaudio/rir/wall.h

@@ -19,16 +19,15 @@ struct Wall {
   const torch::Tensor origin;
   const torch::Tensor normal;
   const torch::Tensor scattering;
-
   const torch::Tensor reflection;
 
   Wall(
       const torch::ArrayRef<scalar_t>& origin,
       const torch::ArrayRef<scalar_t>& normal,
       const torch::Tensor& absorption,
       const torch::Tensor& scattering)
-      : origin(torch::tensor(origin)),
-        normal(torch::tensor(normal)),
+      : origin(torch::tensor(origin).to(scattering.dtype())),
+        normal(torch::tensor(normal).to(scattering.dtype())),
         scattering(scattering),
         reflection(1. - absorption) {}
 };
@@ -137,7 +136,6 @@ std::tuple<torch::Tensor, int, scalar_t> find_collision_wall(
   for (unsigned int i = 0; i < 3; ++i) {
     auto dir0 = SCALAR(direction[i]);
     auto abs_dir0 = std::abs(dir0);
-
     // If the ray is almost parallel to a plane, then we delegate the
     // computation to the other planes.
     if (abs_dir0 < EPS) {
@@ -148,6 +146,10 @@ std::tuple<torch::Tensor, int, scalar_t> find_collision_wall(
     scalar_t distance = (dir0 < 0.)
         ? SCALAR(origin[i]) // Going towards origin
         : SCALAR(room[i] - origin[i]); // Going away from origin
+    // sometimes origin is slightly outside of room
+    if (distance < 0) {
+      distance = 0.;
+    }
     auto ratio = distance / abs_dir0;
     int i_increment = dir0 > 0.;
 

src/torchaudio/prototype/functional/__init__.py

@@ -7,7 +7,7 @@
     oscillator_bank,
     sinc_impulse_response,
 )
-from ._rir import simulate_rir_ism
+from ._rir import ray_tracing, simulate_rir_ism
 from .functional import barkscale_fbanks, chroma_filterbank
 
 
@@ -20,6 +20,7 @@
     "filter_waveform",
     "frequency_impulse_response",
     "oscillator_bank",
+    "ray_tracing",
     "sinc_impulse_response",
     "simulate_rir_ism",
 ]

src/torchaudio/prototype/functional/_rir.py

@@ -133,20 +133,24 @@ def _adjust_coeff(coeffs: Union[float, torch.Tensor], name: str) -> torch.Tensor
     """
     num_walls = 6
     if isinstance(coeffs, float):
+        if coeffs < 0:
+            raise ValueError(f"`{name}` must be non-negative. Found: {coeffs}")
         return torch.full((1, num_walls), coeffs)
     if isinstance(coeffs, Tensor):
+        if torch.any(coeffs < 0):
+            raise ValueError(f"`{name}` must be non-negative. Found: {coeffs}")
         if coeffs.ndim == 1:
             if coeffs.numel() != num_walls:
                 raise ValueError(
-                    f"The shape of `{name}` must be ({num_walls},) when it is a 1D Tensor."
+                    f"The shape of `{name}` must be ({num_walls},) when it is a 1D Tensor. "
                     f"Found the shape {coeffs.shape}."
                 )
             return coeffs.unsqueeze(0)
         if coeffs.ndim == 2:
-            if coeffs.shape != (7, num_walls):
+            if coeffs.shape[1] != num_walls:
                 raise ValueError(
-                    f"The shape of `{name}` must be (7, {num_walls}) when it is a 2D Tensor."
-                    f"Found the shape {coeffs.shape}."
+                    f"The shape of `{name}` must be (NUM_BANDS, {num_walls}) when it "
+                    f"is a 2D Tensor. Found: {coeffs.shape}."
                 )
             return coeffs
     raise TypeError(f"`{name}` must be float or Tensor.")
@@ -169,7 +173,7 @@ def _validate_inputs(
     if not (source.ndim == 1 and source.numel() == 3):
         raise ValueError(f"`source` must be 1D Tensor with 3 elements. Found {source.shape}.")
     if not (mic_array.ndim == 2 and mic_array.shape[1] == 3):
-        raise ValueError(f"mic_array must be a 2D Tensor with shape (num_channels, 3). Found {mic_array.shape}.")
+        raise ValueError(f"`mic_array` must be a 2D Tensor with shape (num_channels, 3). Found {mic_array.shape}.")
 
 
 def simulate_rir_ism(
@@ -270,3 +274,106 @@ def simulate_rir_ism(
             rir = rir[..., :output_length]
 
     return rir
+
+
+def ray_tracing(
+    room: torch.Tensor,
+    source: torch.Tensor,
+    mic_array: torch.Tensor,
+    num_rays: int,
+    absorption: Union[float, torch.Tensor] = 0.0,
+    scattering: Union[float, torch.Tensor] = 0.0,
+    mic_radius: float = 0.5,
+    sound_speed: float = 343.0,
+    energy_thres: float = 1e-7,
+    time_thres: float = 10.0,
+    hist_bin_size: float = 0.004,
+) -> torch.Tensor:
+    r"""Compute energy histogram via ray tracing.
+
+    The implementation is based on *pyroomacoustics* :cite:`scheibler2018pyroomacoustics`.
+
+    ``num_rays`` rays are casted uniformly in all directions from the source;
+    when a ray intersects a wall, it is reflected and part of its energy is absorbed.
+    It is also scattered (sent directly to the microphone(s)) according to the ``scattering``
+    coefficient.
+    When a ray is close to the microphone, its current energy is recorded in the output
+    histogram for that given time slot.
+
+    .. devices:: CPU
+
+    .. properties:: TorchScript
+
+    Args:
+        room (torch.Tensor): Room coordinates. The shape of `room` must be `(3,)` which represents
+            three dimensions of the room.
+        source (torch.Tensor): Sound source coordinates. Tensor with dimensions `(3,)`.
+        mic_array (torch.Tensor): Microphone coordinates. Tensor with dimensions `(channel, 3)`.
+        absorption (float or torch.Tensor, optional): The absorption coefficients of wall materials.
+            (Default: ``0.0``).
+            If the type is ``float``, the absorption coefficient is identical to all walls and
+            all frequencies.
+            If ``absorption`` is a 1D Tensor, the shape must be `(6,)`, representing absorption
+            coefficients of ``"west"``, ``"east"``, ``"south"``, ``"north"``, ``"floor"``, and
+            ``"ceiling"``, respectively.
+            If ``absorption`` is a 2D Tensor, the shape must be  `(num_bands, 6)`.
+            ``num_bands`` is the number of frequency bands (usually 7).
+        scattering(float or torch.Tensor, optional): The scattering coefficients of wall materials. (Default: ``0.0``)
+            The shape and type of this parameter is the same as for ``absorption``.
+        mic_radius(float, optional): The radius of the microphone in meters. (Default: 0.5)
+        sound_speed (float, optional): The speed of sound in meters per second. (Default: ``343.0``)
+        energy_thres (float, optional): The energy level below which we stop tracing a ray. (Default: ``1e-7``)
+            The initial energy of each ray is ``2 / num_rays``.
+        time_thres (float, optional): The maximal duration for which rays are traced. (Unit: seconds) (Default: 10.0)
+        hist_bin_size (float, optional): The size of each bin in the output histogram. (Unit: seconds) (Default: 0.004)
+
+    Returns:
+        (torch.Tensor): The 3D histogram(s) where the energy of the traced ray is recorded.
+            Each bin corresponds to a given time slot.
+            The shape is `(channel, num_bands, num_bins)`, where
+            ``num_bins = ceil(time_thres / hist_bin_size)``.
+            If both ``absorption`` and ``scattering`` are floats, then ``num_bands == 1``.
+    """
+    if time_thres < hist_bin_size:
+        raise ValueError(
+            "`time_thres` must be greater than `hist_bin_size`. "
+            f"Found: hist_bin_size={hist_bin_size}, time_thres={time_thres}."
+        )
+
+    if room.dtype != source.dtype or source.dtype != mic_array.dtype:
+        raise ValueError(
+            "dtype of `room`, `source` and `mic_array` must match. "
+            f"Found: `room` ({room.dtype}), `source` ({source.dtype}) and "
+            f"`mic_array` ({mic_array.dtype})"
+        )
+
+    _validate_inputs(room, source, mic_array)
+    absorption = _adjust_coeff(absorption, "absorption").to(room.dtype)
+    scattering = _adjust_coeff(scattering, "scattering").to(room.dtype)
+
+    # Bring absorption and scattering to the same shape
+    if absorption.shape[0] == 1 and scattering.shape[0] > 1:
+        absorption = absorption.expand(scattering.shape)
+    if scattering.shape[0] == 1 and absorption.shape[0] > 1:
+        scattering = scattering.expand(absorption.shape)
+    if absorption.shape != scattering.shape:
+        raise ValueError(
+            "`absorption` and `scattering` must be broadcastable to the same number of bands and walls. "
+            f"Inferred shapes absorption={absorption.shape} and scattering={scattering.shape}"
+        )
+
+    histograms = torch.ops.torchaudio.ray_tracing(
+        room,
+        source,
+        mic_array,
+        num_rays,
+        absorption,
+        scattering,
+        mic_radius,
+        sound_speed,
+        energy_thres,
+        time_thres,
+        hist_bin_size,
+    )
+
+    return histograms

test/cpp/rir/wall_collision.cpp

@@ -3,22 +3,24 @@
 
 using namespace torchaudio::rir;
 
+using DTYPE = double;
+
 struct CollisionTestParam {
   // Input
   torch::Tensor origin;
   torch::Tensor direction;
   // Expected
   torch::Tensor hit_point;
   int next_wall_index;
-  float hit_distance;
+  DTYPE hit_distance;
 };
 
 CollisionTestParam par(
-    torch::ArrayRef<float> origin,
-    torch::ArrayRef<float> direction,
-    torch::ArrayRef<float> hit_point,
+    torch::ArrayRef<DTYPE> origin,
+    torch::ArrayRef<DTYPE> direction,
+    torch::ArrayRef<DTYPE> hit_point,
     int next_wall_index,
-    float hit_distance) {
+    DTYPE hit_distance) {
   auto dir = torch::tensor(direction);
   return {
       torch::tensor(origin),
@@ -50,18 +52,22 @@ TEST_P(Simple3DRoomCollisionTest, CollisionTest3D) {
 
   auto param = GetParam();
   auto [hit_point, next_wall_index, hit_distance] =
-      find_collision_wall<float>(room, param.origin, param.direction);
+      find_collision_wall<DTYPE>(room, param.origin, param.direction);
 
   EXPECT_EQ(param.next_wall_index, next_wall_index);
   EXPECT_FLOAT_EQ(param.hit_distance, hit_distance);
-  EXPECT_TRUE(torch::allclose(
-      param.hit_point, hit_point, /*rtol*/ 1e-05, /*atol*/ 1e-07));
+  EXPECT_NEAR(
+      param.hit_point[0].item<DTYPE>(), hit_point[0].item<DTYPE>(), 1e-5);
+  EXPECT_NEAR(
+      param.hit_point[1].item<DTYPE>(), hit_point[1].item<DTYPE>(), 1e-5);
+  EXPECT_NEAR(
+      param.hit_point[2].item<DTYPE>(), hit_point[2].item<DTYPE>(), 1e-5);
 }
 
 #define ISQRT2 0.70710678118
 
 INSTANTIATE_TEST_CASE_P(
-    Collision3DTests,
+    BasicCollisionTests,
     Simple3DRoomCollisionTest,
     ::testing::Values(
         // From 0
@@ -100,3 +106,13 @@ INSTANTIATE_TEST_CASE_P(
         par({.5, .5, 1}, {0.0, -1., -1.}, {.5, .0, .5}, 2, ISQRT2),
         par({.5, .5, 1}, {0.0, 1.0, -1.}, {.5, 1., .5}, 3, ISQRT2),
         par({.5, .5, 1}, {0.0, 0.0, -1.}, {.5, .5, .0}, 4, 1.0)));
+
+INSTANTIATE_TEST_CASE_P(
+    CornerCollisionTest,
+    Simple3DRoomCollisionTest,
+    ::testing::Values(
+        par({1, 1, 0}, {1., 1., 0.}, {1., 1., 0.}, 1, 0.0),
+        par({1, 1, 0}, {-1., 1., 0.}, {1., 1., 0.}, 3, 0.0),
+        par({1, 1, 1}, {1., 1., 1.}, {1., 1., 1.}, 1, 0.0),
+        par({1, 1, 1}, {-1., 1., 1.}, {1., 1., 1.}, 3, 0.0),
+        par({1, 1, 1}, {-1., -1., 1.}, {1., 1., 1.}, 5, 0.0)));