Applied contiguous in decode_* ops

test/test_image.py

@@ -94,6 +94,8 @@ def test_decode_jpeg(img_path, pil_mode, mode):
     data = read_file(img_path)
     img_ljpeg = decode_image(data, mode=mode)
 
+    assert img_ljpeg.is_contiguous()
+
     # Permit a small variation on pixel values to account for implementation
     # differences between Pillow and LibJPEG.
     abs_mean_diff = (img_ljpeg.type(torch.float32) - img_pil).abs().mean().item()
@@ -173,6 +175,8 @@ def test_decode_png(img_path, pil_mode, mode):
         data = read_file(img_path)
         img_lpng = decode_image(data, mode=mode)
 
+    assert img_lpng.is_contiguous()
+
     tol = 0 if pil_mode is None else 1
 
     if PILLOW_VERSION >= (8, 3) and pil_mode == "LA":

torchvision/csrc/io/image/cpu/decode_jpeg.cpp

@@ -148,7 +148,7 @@ torch::Tensor decode_jpeg(const torch::Tensor& data, ImageReadMode mode) {
 
   jpeg_finish_decompress(&cinfo);
   jpeg_destroy_decompress(&cinfo);
-  return tensor.permute({2, 0, 1});
+  return tensor.permute({2, 0, 1}).contiguous();
 }
 
 #endif

torchvision/csrc/io/image/cpu/decode_png.cpp

@@ -224,7 +224,7 @@ torch::Tensor decode_png(
     }
   }
   png_destroy_read_struct(&png_ptr, &info_ptr, nullptr);
-  return tensor.permute({2, 0, 1});
+  return tensor.permute({2, 0, 1}).contiguous();
 }
 #endif
 

torchvision/io/image.py

@@ -59,7 +59,7 @@ def write_file(filename: str, data: torch.Tensor) -> None:
 
 def decode_png(input: torch.Tensor, mode: ImageReadMode = ImageReadMode.UNCHANGED) -> torch.Tensor:
     """
-    Decodes a PNG image into a 3 dimensional RGB or grayscale Tensor.
+    Decodes a PNG image into a 3 dimensional RGB or grayscale contiguous Tensor.
     Optionally converts the image to the desired format.
     The values of the output tensor are uint8 in [0, 255].
 
@@ -117,7 +117,7 @@ def decode_jpeg(
     input: torch.Tensor, mode: ImageReadMode = ImageReadMode.UNCHANGED, device: str = "cpu"
 ) -> torch.Tensor:
     """
-    Decodes a JPEG image into a 3 dimensional RGB or grayscale Tensor.
+    Decodes a JPEG image into a 3 dimensional RGB or grayscale contiguous Tensor.
     Optionally converts the image to the desired format.
     The values of the output tensor are uint8 between 0 and 255.
 
@@ -185,7 +185,7 @@ def write_jpeg(input: torch.Tensor, filename: str, quality: int = 75):
 def decode_image(input: torch.Tensor, mode: ImageReadMode = ImageReadMode.UNCHANGED) -> torch.Tensor:
     """
     Detects whether an image is a JPEG or PNG and performs the appropriate
-    operation to decode the image into a 3 dimensional RGB or grayscale Tensor.
+    operation to decode the image into a 3 dimensional RGB or grayscale contiguous Tensor.
 
     Optionally converts the image to the desired format.
     The values of the output tensor are uint8 in [0, 255].
@@ -207,7 +207,7 @@ def decode_image(input: torch.Tensor, mode: ImageReadMode = ImageReadMode.UNCHAN
 
 def read_image(path: str, mode: ImageReadMode = ImageReadMode.UNCHANGED) -> torch.Tensor:
     """
-    Reads a JPEG or PNG image into a 3 dimensional RGB or grayscale Tensor.
+    Reads a JPEG or PNG image into a 3 dimensional RGB or grayscale contiguous Tensor.
     Optionally converts the image to the desired format.
     The values of the output tensor are uint8 in [0, 255].