PR: Add libpng and libjpeg-turbo requirement into conda recipe (#2301)

* Add libpng requirement into conda recipe

* Try to install libjpeg-turbo

* Add PNG reading capabilities

* Remove newline

* Add image extension to compilation instructions

* Include png functions as part of the main library

* Update CMakeLists

* Detect if building on conda-build

* Debug

* More debug messages

* Print globbed libreries

* Print globbed libreries

* Point to correct PNG path

* Remove libJPEG preventively

* Debug extension loading

* Link libpng explicitly

* Link with PNG

* Add PNG reading capabilities

* Add libpng requirement into conda recipe

* Try to install libjpeg-turbo

* Remove newline

* Add image extension to compilation instructions

* Include png functions as part of the main library

* Update CMakeLists

* Detect if building on conda-build

* Debug

* More debug messages

* Print globbed libreries

* Print globbed libreries

* Point to correct PNG path

* Remove libJPEG preventively

* Debug extension loading

* Link libpng explicitly

* Link with PNG

* Install libpng on conda-based wheel distributions

* Add -y flag

* Add -y flag to yum

* Locate LibPNG on windows conda

* Remove empty else

* Copy libpng16.so

* Copy dylib on Mac

* Improve check on Windows

* Try to install ninja using conda on windows

* Use libpng on Windows

* Package lib on windows wheel

* Point library to the correct place

* Include binaries as part of wheel

* Copy libpng.so on linux

* Look for png.h on Windows when using conda-build

* Do not skip png tests on Mac/Win

* Restore libjpeg-turbo

* Install jpeg-turbo on wheel distributions

* Install libjpeg-turbo from conda-forge on wheel distributions

* Do not pull av on conda-build

* Add pillow disclaimer

* Vendors libjpeg-turbo 2.0.4

* Merge JPEG work

* Remove submodules

* Regenerate circle config

* Fix style issues

* Fix C++ style issues

* More style corrections

* Add JPEG-turbo to linking libraries

* More style corrections

* More style corrections

* More style corrections

* Install libjpeg-turbo-devel

* Install libturbo-jpeg on typing pipeline

* Update Circle template

* Windows and Unix turbojpeg have the same linking name

* Install turbojpeg-devel instead of libjpeg-turbo

* Copy TurboJPEG binaries to wheel

* Move test image

* Move back test image

* Update JPEG test path

* Remove dot from extension

* Move image functions to extension

* Use stdout arg in subprocess

* Disable image extension if libpng or turbojpeg are not found

* Append libpng stdout

* Prevent list appending on lists

* Minor path correction

* Minor error correction

* Add linking flags

* Style issues correction

* Address minor review corrections

* Refactor library search

* Restore access index

* Fix JPEG tests

* Update libpng version in Travis

* Add -y flag

* Remove dot

* Update libpng using apt

* Check libpng version

* Change libturbojpeg binary

* Update import

* Change call

* Restore av in conda recipe

* Minor error correction

* Remove unused comment in travis.yml

* Update README

* Fix missing links

* Remove fixes for 16.04
Co-authored-by: Ryad ZENINE <r.zenine@gmail.com>

.circleci/config.yml

@@ -107,6 +107,8 @@ jobs:
       - checkout
       - run:
           command: |
+            sudo apt-get update -y
+            sudo apt install -y libturbojpeg-dev
             pip install --user --progress-bar off numpy mypy
             pip install --user --progress-bar off --pre torch -f https://download.pytorch.org/whl/nightly/cpu/torch_nightly.html
             pip install --user --progress-bar off --editable .

.circleci/config.yml.in

@@ -107,6 +107,8 @@ jobs:
       - checkout
       - run:
           command: |
+            sudo apt-get update -y
+            sudo apt install -y libturbojpeg-dev
             pip install --user --progress-bar off numpy mypy
             pip install --user --progress-bar off --pre torch -f https://download.pytorch.org/whl/nightly/cpu/torch_nightly.html
             pip install --user --progress-bar off --editable .

.travis.yml

@@ -13,6 +13,7 @@ jobs:
 
 before_install:
   - sudo apt-get update
+  - sudo apt-get install -y libpng-dev libturbojpeg-dev
   - wget https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh -O miniconda.sh;
   - bash miniconda.sh -b -p $HOME/miniconda
   - export PATH="$HOME/miniconda/bin:$PATH"

CMakeLists.txt

@@ -11,22 +11,28 @@ if(WITH_CUDA)
 endif()
 
 find_package(Python3 COMPONENTS Development)
+
 find_package(Torch REQUIRED)
+find_package(PNG REQUIRED)
+
 
 file(GLOB HEADERS torchvision/csrc/*.h)
-file(GLOB OPERATOR_SOURCES torchvision/csrc/cpu/*.h torchvision/csrc/cpu/*.cpp torchvision/csrc/*.cpp)
+file(GLOB IMAGE_HEADERS torchvision/csrc/cpu/image/*.h)
+file(GLOB IMAGE_SOURCES torchvision/csrc/cpu/image/*.cpp)
+file(GLOB OPERATOR_SOURCES torchvision/csrc/cpu/*.h torchvision/csrc/cpu/*.cpp ${IMAGE_HEADERS} ${IMAGE_SOURCES} ${HEADERS} torchvision/csrc/*.cpp)
 if(WITH_CUDA)
   file(GLOB OPERATOR_SOURCES ${OPERATOR_SOURCES} torchvision/csrc/cuda/*.h torchvision/csrc/cuda/*.cu)
 endif()
 file(GLOB MODELS_HEADERS torchvision/csrc/models/*.h)
 file(GLOB MODELS_SOURCES torchvision/csrc/models/*.h torchvision/csrc/models/*.cpp)
 
-add_library(${PROJECT_NAME} SHARED ${MODELS_SOURCES} ${OPERATOR_SOURCES})
-target_link_libraries(${PROJECT_NAME} PRIVATE ${TORCH_LIBRARIES} Python3::Python)
+add_library(${PROJECT_NAME} SHARED ${MODELS_SOURCES} ${OPERATOR_SOURCES} ${IMAGE_SOURCES})
+target_link_libraries(${PROJECT_NAME} PRIVATE ${TORCH_LIBRARIES} ${PNG_LIBRARY} Python3::Python)
+# target_link_libraries(${PROJECT_NAME} PRIVATE ${PNG_LIBRARY} Python3::Python)
 set_target_properties(${PROJECT_NAME} PROPERTIES EXPORT_NAME TorchVision)
 
 target_include_directories(${PROJECT_NAME} INTERFACE
-  $<BUILD_INTERFACE:${HEADERS}>
+  $<BUILD_INTERFACE:${HEADERS}:${PNG_INCLUDE_DIR}>
   $<INSTALL_INTERFACE:${CMAKE_INSTALL_INCLUDEDIR}>)
 
 include(GNUInstallDirs)

README.rst

@@ -78,6 +78,15 @@ Torchvision currently supports the following image backends:
 
 * `accimage`_ - if installed can be activated by calling :code:`torchvision.set_image_backend('accimage')`
 
+* `libpng`_ - can be installed via conda :code:`conda install libpng` or any of the package managers for debian-based and RHEL-based Linux distributions.
+
+* `libturbojpeg`_ - blazing speed, fast JPEG image loading. Can be installed from conda-forge :code:`conda install libjpeg-turbo -c conda-forge`.
+
+**Notes:** ``libpng`` and ``libturbojpeg`` must be available at compilation time in order to be available. Also, most linux distributions distinguish between
+``libturbojpeg`` and ``libjpeg-turbo``, where the former should be installed instead of the latter one.
+
+.. _libpng : http://www.libpng.org/pub/png/libpng.html
+.. _libturbojpeg: https://github.com/libjpeg-turbo/libjpeg-turbo
 .. _Pillow : https://python-pillow.org/
 .. _Pillow-SIMD : https://github.com/uploadcare/pillow-simd
 .. _accimage: https://github.com/pytorch/accimage

packaging/build_wheel.sh

@@ -10,6 +10,30 @@ setup_wheel_python
 pip_install numpy pyyaml future ninja
 setup_pip_pytorch_version
 python setup.py clean
+
+# Copy binaries to be included in the wheel distribution
+if [[ "$(uname)" == Darwin || "$OSTYPE" == "msys" ]]; then
+    python_exec="$(which python)"
+    bin_path=$(dirname $python_exec)
+    env_path=$(dirname $bin_path)
+    if [[ "$(uname)" == Darwin ]]; then
+        # Include LibPNG
+        cp "$env_path/lib/libpng16.dylib" torchvision
+        # Include TurboJPEG
+        cp "$env_path/lib/libturbojpeg.dylib" torchvision
+    else
+        # Include libPNG
+        cp "$bin_path/Library/lib/libpng.lib" torchvision
+        # Include TurboJPEG
+        cp "$bin_path/Library/lib/turbojpeg.lib" torchvision
+    fi
+else
+    # Include LibPNG
+    cp "/usr/lib64/libpng.so" torchvision
+    # Include TurboJPEG
+    cp "/usr/lib64/libturbojpeg.so" torchvision
+fi
+
 if [[ "$OSTYPE" == "msys" ]]; then
     IS_WHEEL=1 "$script_dir/windows/internal/vc_env_helper.bat" python setup.py bdist_wheel
 else

packaging/pkg_helpers.bash

@@ -170,7 +170,13 @@ setup_wheel_python() {
     conda env remove -n "env$PYTHON_VERSION" || true
     conda create -yn "env$PYTHON_VERSION" python="$PYTHON_VERSION"
     conda activate "env$PYTHON_VERSION"
+    # Install libPNG from Anaconda (defaults)
+    conda install libpng -y
+    # Install libJPEG-turbo from conda-forge
+    conda install -y libjpeg-turbo -c conda-forge
   else
+    # Install native CentOS libPNG, libJPEG-turbo
+    yum install -y libpng-devel turbojpeg-devel
     case "$PYTHON_VERSION" in
       2.7)
         if [[ -n "$UNICODE_ABI" ]]; then

packaging/torchvision/conda_build_config.yaml

+channel_sources:
+  - defaults,conda-forge
+
 blas_impl:
   - mkl                        # [x86_64]
 c_compiler:

packaging/torchvision/meta.yaml

@@ -8,6 +8,8 @@ source:
 requirements:
   build:
     - {{ compiler('c') }} # [win]
+    - libpng
+    - libjpeg-turbo
 
   host:
     - python
@@ -18,6 +20,10 @@ requirements:
 
   run:
     - python
+    - libpng
+    - libjpeg-turbo
+    # Pillow introduces unwanted conflicts with libjpeg-turbo, as it depends on jpeg
+    # The fix depends on https://github.com/conda-forge/conda-forge.github.io/issues/673
     - pillow >=4.1.1
     - numpy >=1.11
     {{ environ.get('CONDA_PYTORCH_CONSTRAINT') }}

setup.py

@@ -3,7 +3,7 @@ import io
 import re
 import sys
 from setuptools import setup, find_packages
-from pkg_resources import get_distribution, DistributionNotFound
+from pkg_resources import parse_version, get_distribution, DistributionNotFound
 import subprocess
 import distutils.command.clean
 import distutils.spawn
@@ -76,7 +76,76 @@ pillow_req = 'pillow-simd' if get_dist('pillow-simd') is not None else 'pillow'
 requirements.append(pillow_req + pillow_ver)
 
 
+def find_library(name, vision_include):
+    this_dir = os.path.dirname(os.path.abspath(__file__))
+    build_prefix = os.environ.get('BUILD_PREFIX', None)
+    is_conda_build = build_prefix is not None
+
+    library_found = False
+    conda_installed = False
+    lib_folder = None
+    include_folder = None
+    library_header = '{0}.h'.format(name)
+
+    print('Running build on conda-build: {0}'.format(is_conda_build))
+    if is_conda_build:
+        # Add conda headers/libraries
+        if os.name == 'nt':
+            build_prefix = os.path.join(build_prefix, 'Library')
+        include_folder = os.path.join(build_prefix, 'include')
+        lib_folder = os.path.join(build_prefix, 'lib')
+        library_header_path = os.path.join(
+            include_folder, library_header)
+        library_found = os.path.isfile(library_header_path)
+        conda_installed = library_found
+    else:
+        # Check if using Anaconda to produce wheels
+        conda = distutils.spawn.find_executable('conda')
+        is_conda = conda is not None
+        print('Running build on conda: {0}'.format(is_conda))
+        if is_conda:
+            python_executable = sys.executable
+            py_folder = os.path.dirname(python_executable)
+            if os.name == 'nt':
+                env_path = os.path.join(py_folder, 'Library')
+            else:
+                env_path = os.path.dirname(py_folder)
+            lib_folder = os.path.join(env_path, 'lib')
+            include_folder = os.path.join(env_path, 'include')
+            library_header_path = os.path.join(
+                include_folder, library_header)
+            library_found = os.path.isfile(library_header_path)
+            conda_installed = library_found
+
+    # Try to locate turbojpeg in Linux standard paths
+    if not library_found:
+        if sys.platform == 'linux':
+            library_found = os.path.exists('/usr/include/{0}'.format(
+                library_header))
+            library_found = library_found or os.path.exists(
+                '/usr/local/include/{0}'.format(library_header))
+        else:
+            # Lookup in TORCHVISION_INCLUDE or in the package file
+            package_path = [os.path.join(this_dir, 'torchvision')]
+            for folder in vision_include + package_path:
+                candidate_path = os.path.join(folder, library_header)
+                library_found = os.path.exists(candidate_path)
+                if library_found:
+                    break
+
+    return library_found, conda_installed, include_folder, lib_folder
+
+
 def get_extensions():
+    vision_include = os.environ.get('TORCHVISION_INCLUDE', None)
+    vision_library = os.environ.get('TORCHVISION_LIBRARY', None)
+    vision_include = (vision_include.split(os.pathsep)
+                      if vision_include is not None else [])
+    vision_library = (vision_library.split(os.pathsep)
+                      if vision_library is not None else [])
+    include_dirs = vision_include
+    library_dirs = vision_library
+
     this_dir = os.path.dirname(os.path.abspath(__file__))
     extensions_dir = os.path.join(this_dir, 'torchvision', 'csrc')
 
@@ -149,13 +218,14 @@ def get_extensions():
 
     sources = [os.path.join(extensions_dir, s) for s in sources]
 
-    include_dirs = [extensions_dir]
+    include_dirs += [extensions_dir]
 
     ext_modules = [
         extension(
             'torchvision._C',
             sources,
             include_dirs=include_dirs,
+            library_dirs=library_dirs,
             define_macros=define_macros,
             extra_compile_args=extra_compile_args,
         )
@@ -171,6 +241,65 @@ def get_extensions():
             )
         )
 
+    # Image reading extension
+    image_macros = []
+    image_include = [extensions_dir]
+    image_library = []
+    image_link_flags = []
+
+    # Locating libPNG
+    libpng = distutils.spawn.find_executable('libpng-config')
+    png_found = libpng is not None
+    image_macros += [('PNG_FOUND', str(int(png_found)))]
+    print('PNG found: {0}'.format(png_found))
+    if png_found:
+        png_version = subprocess.run([libpng, '--version'],
+                                     stdout=subprocess.PIPE)
+        png_version = png_version.stdout.strip().decode('utf-8')
+        print('libpng version: {0}'.format(png_version))
+        png_version = parse_version(png_version)
+        if png_version >= parse_version("1.6.0"):
+            print('Building torchvision with PNG image support')
+            png_lib = subprocess.run([libpng, '--libdir'],
+                                     stdout=subprocess.PIPE)
+            png_include = subprocess.run([libpng, '--I_opts'],
+                                         stdout=subprocess.PIPE)
+            image_library += [png_lib.stdout.strip().decode('utf-8')]
+            image_include += [png_include.stdout.strip().decode('utf-8')]
+            image_link_flags.append('png' if os.name != 'nt' else 'libpng')
+        else:
+            print('libpng installed version is less than 1.6.0, '
+                  'disabling PNG support')
+            png_found = False
+
+    # Locating libjpegturbo
+    turbojpeg_info = find_library('turbojpeg', vision_include)
+    (turbojpeg_found, conda_installed,
+     turbo_include_folder, turbo_lib_folder) = turbojpeg_info
+
+    image_macros += [('JPEG_FOUND', str(int(turbojpeg_found)))]
+    print('turboJPEG found: {0}'.format(turbojpeg_found))
+    if turbojpeg_found:
+        print('Building torchvision with JPEG image support')
+        image_link_flags.append('turbojpeg')
+        if conda_installed:
+            image_library += [turbo_lib_folder]
+            image_include += [turbo_include_folder]
+
+    image_path = os.path.join(extensions_dir, 'cpu', 'image')
+    image_src = glob.glob(os.path.join(image_path, '*.cpp'))
+
+    if png_found or turbojpeg_found:
+        ext_modules.append(extension(
+            'torchvision.image',
+            image_src,
+            include_dirs=include_dirs + [image_path] + image_include,
+            library_dirs=library_dirs + image_library,
+            define_macros=image_macros,
+            libraries=image_link_flags,
+            extra_compile_args=extra_compile_args
+        ))
+
     ffmpeg_exe = distutils.spawn.find_executable('ffmpeg')
     has_ffmpeg = ffmpeg_exe is not None
 
@@ -243,7 +372,9 @@ setup(
 
     # Package info
     packages=find_packages(exclude=('test',)),
-
+    package_data={
+        package_name: ['*.lib', '*.dylib', '*.so']
+    },
     zip_safe=False,
     install_requires=requirements,
     extras_require={

test/test_image.py

+import os
+import unittest
+import sys
+
+import torch
+import torchvision
+from PIL import Image
+from torchvision.io.image import read_png, decode_png, read_jpeg, decode_jpeg
+import numpy as np
+
+IMAGE_ROOT = os.path.join(os.path.dirname(os.path.abspath(__file__)), "assets")
+IMAGE_DIR = os.path.join(IMAGE_ROOT, "fakedata", "imagefolder")
+
+
+def get_images(directory, img_ext):
+    assert os.path.isdir(directory)
+    for root, _, files in os.walk(directory):
+        for fl in files:
+            _, ext = os.path.splitext(fl)
+            if ext == img_ext:
+                yield os.path.join(root, fl)
+
+
+class ImageTester(unittest.TestCase):
+    def test_read_jpeg(self):
+        for img_path in get_images(IMAGE_ROOT, "jpg"):
+            img_pil = torch.from_numpy(np.array(Image.open(img_path)))
+            img_ljpeg = read_jpeg(img_path)
+
+            norm = img_ljpeg.shape[0] * img_ljpeg.shape[1] * img_ljpeg.shape[2] * 255
+            err = torch.abs(img_ljpeg.flatten().float() - img_pil.flatten().float()).sum().float() / (norm)
+            self.assertLessEqual(err, 1e-2)
+
+    def test_decode_jpeg(self):
+        for img_path in get_images(IMAGE_ROOT, "jpg"):
+            img_pil = torch.from_numpy(np.array(Image.open(img_path)))
+            size = os.path.getsize(img_path)
+            img_ljpeg = decode_jpeg(torch.from_file(img_path, dtype=torch.uint8, size=size))
+
+            norm = img_ljpeg.shape[0] * img_ljpeg.shape[1] * img_ljpeg.shape[2] * 255
+            err = torch.abs(img_ljpeg.flatten().float() - img_pil.flatten().float()).sum().float() / (norm)
+
+            self.assertLessEqual(err, 1e-2)
+
+        with self.assertRaisesRegex(ValueError, "Expected a non empty 1-dimensional tensor."):
+            decode_jpeg(torch.empty((100, 1), dtype=torch.uint8))
+
+        with self.assertRaisesRegex(ValueError, "Expected a torch.uint8 tensor."):
+            decode_jpeg(torch.empty((100, ), dtype=torch.float16))
+
+        with self.assertRaisesRegex(RuntimeError, "Error while reading jpeg headers"):
+            decode_jpeg(torch.empty((100), dtype=torch.uint8))
+
+    def test_read_png(self):
+        for img_path in get_images(IMAGE_DIR, "png"):
+            img_pil = torch.from_numpy(np.array(Image.open(img_path)))
+            img_lpng = read_png(img_path)
+            self.assertEqual(img_lpng, img_pil)
+
+    def test_decode_png(self):
+        for img_path in get_images(IMAGE_DIR, "png"):
+            img_pil = torch.from_numpy(np.array(Image.open(img_path)))
+            size = os.path.getsize(img_path)
+            img_lpng = decode_png(torch.from_file(img_path, dtype=torch.uint8, size=size))
+            self.assertEqual(img_lpng, img_pil)
+
+            self.assertEqual(decode_png(torch.empty()), torch.empty())
+            self.assertEqual(decode_png(torch.randint(3, 5, (300,))), torch.empty())
+
+
+if __name__ == '__main__':
+    unittest.main()

torchvision/csrc/cpu/image/image.cpp

+
+#include "image.h"
+#include <ATen/ATen.h>
+#include <Python.h>
+
+// If we are in a Windows environment, we need to define
+// initialization functions for the _custom_ops extension
+#ifdef _WIN32
+#if PY_MAJOR_VERSION < 3
+PyMODINIT_FUNC init_image(void) {
+  // No need to do anything.
+  return NULL;
+}
+#else
+PyMODINIT_FUNC PyInit_image(void) {
+  // No need to do anything.
+  return NULL;
+}
+#endif
+#endif
+
+static auto registry = torch::RegisterOperators()
+                           .op("image::decode_png", &decodePNG)
+                           .op("image::decode_jpeg", &decodeJPEG);

torchvision/csrc/cpu/image/image.h

+
+#pragma once
+
+#include <torch/script.h>
+#include <torch/torch.h>
+#include "readjpeg_cpu.h"
+#include "readpng_cpu.h"

torchvision/csrc/cpu/image/readjpeg_cpu.cpp

+#include "readjpeg_cpu.h"
+
+#include <ATen/ATen.h>
+#include <setjmp.h>
+#include <string>
+
+#if !JPEG_FOUND
+
+torch::Tensor decodeJPEG(const torch::Tensor& data) {
+  AT_ERROR("decodeJPEG: torchvision not compiled with turboJPEG support");
+}
+
+#else
+#include <turbojpeg.h>
+
+torch::Tensor decodeJPEG(const torch::Tensor& data) {
+  tjhandle tjInstance = tjInitDecompress();
+  if (tjInstance == NULL) {
+    TORCH_CHECK(false, "libjpeg-turbo decompression initialization failed.");
+  }
+
+  auto datap = data.accessor<unsigned char, 1>().data();
+
+  int width, height;
+
+  if (tjDecompressHeader(tjInstance, datap, data.numel(), &width, &height) <
+      0) {
+    tjDestroy(tjInstance);
+    TORCH_CHECK(false, "Error while reading jpeg headers");
+  }
+  auto tensor =
+      torch::empty({int64_t(height), int64_t(width), int64_t(3)}, torch::kU8);
+
+  auto ptr = tensor.accessor<uint8_t, 3>().data();
+
+  int pixelFormat = TJPF_RGB;
+
+  auto ret = tjDecompress2(
+      tjInstance,
+      datap,
+      data.numel(),
+      ptr,
+      width,
+      0,
+      height,
+      pixelFormat,
+      NULL);
+  if (ret != 0) {
+    tjDestroy(tjInstance);
+    TORCH_CHECK(false, "decompressing JPEG image");
+  }
+
+  return tensor;
+}
+
+#endif // JPEG_FOUND

torchvision/csrc/cpu/image/readjpeg_cpu.h

+#pragma once
+
+#include <torch/torch.h>
+
+torch::Tensor decodeJPEG(const torch::Tensor& data);

torchvision/csrc/cpu/image/readpng_cpu.cpp

+#include "readpng_cpu.h"
+
+#include <ATen/ATen.h>
+#include <setjmp.h>
+#include <string>
+
+#if !PNG_FOUND
+torch::Tensor decodePNG(const torch::Tensor& data) {
+  AT_ERROR("decodePNG: torchvision not compiled with libPNG support");
+}
+#else
+#include <png.h>
+
+torch::Tensor decodePNG(const torch::Tensor& data) {
+  auto png_ptr =
+      png_create_read_struct(PNG_LIBPNG_VER_STRING, nullptr, nullptr, nullptr);
+  TORCH_CHECK(png_ptr, "libpng read structure allocation failed!")
+  auto info_ptr = png_create_info_struct(png_ptr);
+  if (!info_ptr) {
+    png_destroy_read_struct(&png_ptr, nullptr, nullptr);
+    // Seems redundant with the if statement. done here to avoid leaking memory.
+    TORCH_CHECK(info_ptr, "libpng info structure allocation failed!")
+  }
+
+  auto datap = data.accessor<unsigned char, 1>().data();
+
+  if (setjmp(png_jmpbuf(png_ptr)) != 0) {
+    png_destroy_read_struct(&png_ptr, &info_ptr, nullptr);
+    TORCH_CHECK(false, "Internal error.");
+  }
+  auto is_png = !png_sig_cmp(datap, 0, 8);
+  TORCH_CHECK(is_png, "Content is not png!")
+
+  struct Reader {
+    png_const_bytep ptr;
+  } reader;
+  reader.ptr = png_const_bytep(datap) + 8;
+
+  auto read_callback =
+      [](png_structp png_ptr, png_bytep output, png_size_t bytes) {
+        auto reader = static_cast<Reader*>(png_get_io_ptr(png_ptr));
+        std::copy(reader->ptr, reader->ptr + bytes, output);
+        reader->ptr += bytes;
+      };
+  png_set_sig_bytes(png_ptr, 8);
+  png_set_read_fn(png_ptr, &reader, read_callback);
+  png_read_info(png_ptr, info_ptr);
+
+  png_uint_32 width, height;
+  int bit_depth, color_type;
+  auto retval = png_get_IHDR(
+      png_ptr,
+      info_ptr,
+      &width,
+      &height,
+      &bit_depth,
+      &color_type,
+      nullptr,
+      nullptr,
+      nullptr);
+
+  if (retval != 1) {
+    png_destroy_read_struct(&png_ptr, &info_ptr, nullptr);
+    TORCH_CHECK(retval == 1, "Could read image metadata from content.")
+  }
+  if (color_type != PNG_COLOR_TYPE_RGB) {
+    png_destroy_read_struct(&png_ptr, &info_ptr, nullptr);
+    TORCH_CHECK(
+        color_type == PNG_COLOR_TYPE_RGB, "Non RGB images are not supported.")
+  }
+
+  auto tensor =
+      torch::empty({int64_t(height), int64_t(width), int64_t(3)}, torch::kU8);
+  auto ptr = tensor.accessor<uint8_t, 3>().data();
+  auto bytes = png_get_rowbytes(png_ptr, info_ptr);
+  for (decltype(height) i = 0; i < height; ++i) {
+    png_read_row(png_ptr, ptr, nullptr);
+    ptr += bytes;
+  }
+  png_destroy_read_struct(&png_ptr, &info_ptr, nullptr);
+  return tensor;
+}
+#endif // PNG_FOUND

torchvision/csrc/cpu/image/readpng_cpu.h

+#pragma once
+
+#include <torch/torch.h>
+#include <string>
+
+torch::Tensor decodePNG(const torch::Tensor& data);

torchvision/io/__init__.py

@@ -30,5 +30,5 @@ __all__ = [
     "_read_video_clip_from_memory",
     "_read_video_meta_data",
     "VideoMetaData",
-    "Timebase",
+    "Timebase"
 ]

torchvision/io/image.py

+import torch
+from torch import nn, Tensor
+
+import os
+import os.path as osp
+import importlib
+
+_HAS_IMAGE_OPT = False
+
+try:
+    lib_dir = osp.join(osp.dirname(__file__), "..")
+
+    loader_details = (
+        importlib.machinery.ExtensionFileLoader,
+        importlib.machinery.EXTENSION_SUFFIXES
+    )
+
+    extfinder = importlib.machinery.FileFinder(lib_dir, loader_details)
+    ext_specs = extfinder.find_spec("image")
+    if ext_specs is not None:
+        torch.ops.load_library(ext_specs.origin)
+        _HAS_IMAGE_OPT = True
+except (ImportError, OSError):
+    pass
+
+
+def decode_png(input):
+    # type: (Tensor) -> Tensor
+    """
+    Decodes a PNG image into a 3 dimensional RGB Tensor.
+    The values of the output tensor are uint8 between 0 and 255.
+
+    Arguments:
+        input (Tensor[1]): a one dimensional int8 tensor containing
+    the raw bytes of the PNG image.
+
+    Returns:
+        output (Tensor[image_width, image_height, 3])
+    """
+    if not isinstance(input, torch.Tensor) or len(input) == 0:
+        raise ValueError("Expected a non empty 1-dimensional tensor.")
+
+    if not input.dtype == torch.uint8:
+        raise ValueError("Expected a torch.uint8 tensor.")
+    output = torch.ops.image.decode_png(input)
+    return output
+
+
+def read_png(path):
+    # type: (str) -> Tensor
+    """
+    Reads a PNG image into a 3 dimensional RGB Tensor.
+    The values of the output tensor are uint8 between 0 and 255.
+
+    Arguments:
+        path (str): path of the PNG image.
+
+    Returns:
+        output (Tensor[image_width, image_height, 3])
+    """
+    if not os.path.isfile(path):
+        raise ValueError("Expected a valid file path.")
+
+    size = os.path.getsize(path)
+    if size == 0:
+        raise ValueError("Expected a non empty file.")
+    data = torch.from_file(path, dtype=torch.uint8, size=size)
+    return decode_png(data)
+
+
+def decode_jpeg(input):
+    # type: (Tensor) -> Tensor
+    """
+    Decodes a JPEG image into a 3 dimensional RGB Tensor.
+    The values of the output tensor are uint8 between 0 and 255.
+    Arguments:
+        input (Tensor[1]): a one dimensional int8 tensor containing
+    the raw bytes of the JPEG image.
+    Returns:
+        output (Tensor[image_width, image_height, 3])
+    """
+    if not isinstance(input, torch.Tensor) or len(input) == 0 or input.ndim != 1:
+        raise ValueError("Expected a non empty 1-dimensional tensor.")
+
+    if not input.dtype == torch.uint8:
+        raise ValueError("Expected a torch.uint8 tensor.")
+
+    output = torch.ops.image.decode_jpeg(input)
+    return output
+
+
+def read_jpeg(path):
+    # type: (str) -> Tensor
+    """
+    Reads a JPEG image into a 3 dimensional RGB Tensor.
+    The values of the output tensor are uint8 between 0 and 255.
+    Arguments:
+        path (str): path of the JPEG image.
+    Returns:
+        output (Tensor[image_width, image_height, 3])
+    """
+    if not os.path.isfile(path):
+        raise ValueError("Expected a valid file path.")
+
+    size = os.path.getsize(path)
+    if size == 0:
+        raise ValueError("Expected a non empty file.")
+    data = torch.from_file(path, dtype=torch.uint8, size=size)
+    return decode_jpeg(data)