🚀 The feature

Torchvision'sread_image currently decodes JPEG images at full resolution. However, bothlibjpeg andlibjpeg-turbo support decoding at lower resolutions (1/2, 1/4, 1/8 of the original size).

Introducing asize_hint parameter would allow users to specify an approximate target size, withtorchvision selecting the closest larger available scale factor and downscale the JPEG image during decoding.

Example Usage:

fromtorchvision.io.imageimportdecode_imagetensor=decode_image("image.jpeg",size_hint=(224,224))

Motivation, pitch

• Many ML pipelines process images at fixed sizes (e.g., 224x224 for ImageNet models). Decoding large images only to downscale them later is inefficient.
• This can improve memory usage as we do not need to hold the full-sized image in the memory.
• Pillow provides a similar feature viaImage.draft, allowing for approximate size-based decoding.

Alternatives

• Using Pillow for decoding with downscaling, but torchvision’s native decoder is typically faster than decoding using Pillow and then converting to tensor.
• Decode and then resize, but this is inefficient, see benchmark below.

Additional context

Benchmark

We implemented a proof-of-concept and ran performance tests on decoding a 1920x1080 image into 960x540.
We compared the following:

• Use existingdecode_jpeg and resize after.
• Patchdecode_jpeg to allowlibjpeg /libjpeg-turbo downscaling via thesize_hint parameters.

Benchmark results (1000 iters):

9.91s call     .../test_jpeg.py::test_torchvision_image_load_with_resize_960_5404.00s call     .../test_jpeg.py::test_fastjpeg_image_load_with_size_hint_960_540

~2.5X speed up.

I'm happy to contribute a patch if people consider this useful.