pytorch/vision

Issues

• Pretrained weights for ResNet[18, 34, 50, 101] are incorrect

  #8749 opened a month ago by longyuxi
  3

• Add Points to TVTensors

  #8728 opened 2 months ago by Alexandre-SCHOEPP
  1

• The link of **Multi-view Stereo Correspondence** doesn't exist in the doc

  #8722 opened 2 months ago by hyperkai
  3

• make processing of arbitrary inputs to transforms.v2 public and document it

  #8721 opened 2 months ago by liopeer
  2

• Setting `flickr30k` to `root` argument of `Flickr30k()` gets error

  #8750 opened 2 months ago by hyperkai
  1

• Setting `flickr8k` from `illinois.edu` to `root` argument of `Flickr8k()` doesn't load any images, returning the length `0`

  #8748 opened 2 months ago by hyperkai
  1

• The link of `ImageNet()` doesn't have `ILSVRC2012_img_train.tar` and `ILSVRC2012_img_val.tar`.

  #8745 opened a month ago by hyperkai
  1

• The link of `Flickr30k()` doesn't have a Flickr30k dataset

  #8742 opened a month ago by hyperkai
  1

• `split` argument of `Places365()` should accept `"test"`

  #8741 opened 2 months ago by hyperkai
  1

• Downloading data twice with `download=True` set to `Places365()` should not get error like `MNIST()`

  #8740 opened a month ago by hyperkai
  1

• torchvision.ops DeformConv2d creates RuntimeError: CUDA error: an illegal memory access was encountered on CUDA and not on CPU

  #8744 opened 2 months ago by sohamroy2538
  1

• Support for python 3.13

  #8747 opened 2 months ago by SH2282000
  1

• release for python 3.13

  #8730 opened 2 months ago by dpinol
  1

• MPS Issue : view size is not compatible with input tensor's size and stride

  #8706 opened a month ago by shanalikhan
  7

• `_skip_resize` ignored on detector inferenece

  #8735 opened 2 months ago by talcs
  2

• optests.generate_opcheck_tests

  #8731 opened 2 months ago by goldfishsound
  3

• FasterRcnn conversion to ONNX example fails

  #8734 opened 2 months ago by cholland-nv
  1

• Building libtorchvision.so with nvJPEG

  #8751 opened 2 months ago by srasmussenvl
  1

• The link of LSUN doesn't exist in the doc

  #8738 opened a month ago by hyperkai
  2

• The link of LSUN doesn't exist in the doc

  #8739 opened a month ago by hyperkai
  1

• datasets phototour.py gives errors

  #8732 opened 2 months ago by venkatram-dev
  2

• Conversion of WEBP to grayscale on read

  #8753 opened 2 months ago by DLumi
  2

• MNIST url update

  #8752 opened 2 months ago by rmchurch
  1

• The error messasge of `classes` argument of `LSUN()` should be corrected

  #8756 opened a month ago by hyperkai
  1

• [BUG] Found a bug in GoogLeNet. The order of aux1 and aux2 is incorrectly reversed.

  #8736 opened 2 months ago by pblwk
  0

• nms

  #8729 opened 2 months ago by 5464jmy
  1

• Video classification dataset cache can't be loaded in reference training code

  #8726 opened 2 months ago by Met4physics
  1

• Downloading MNIST always fails on the first mirror due to an SSL certificate error

  #8717 opened 2 months ago by pczarnik
  2

• wrap_dataset_for_transforms_v2 with transforms not working as intended

  #8720 opened 2 months ago by liopeer
  0

• `torchvision.datasets.ImageNet()` race condition when multi-process training launched with `root` containing only *.tar *.tar.gz downloaded files

  #8707 opened 2 months ago by EIFY
  2

• Wrong code in setup.py: it assumes that it is Windows if it's not linux or darwin

  #8716 opened 2 months ago by yurivict
  1

• Suspected memory leak in decode_image when reading webp files

  #8710 opened 2 months ago by hassonofer
  3

• ld: error: unable to find library -llibpng

  #8715 opened 2 months ago by yurivict
  1

• I am using the torchvision-0.13.1+cu113 version, but it seems that it does not have the datapoints package. How can I solve this issue?

  #8714 opened 2 months ago by jiangsu415
  2

• `torchvision.ops.boxes.batched_nms` slow on large box numbers

  #8713 opened 2 months ago by Ghelfi
  0

• Phototour dataset URLs are broken

  #8711 opened 2 months ago by RiddleMa
  1

• `from torchvision.datasets import *` i can't through this way import MovingMNIST

  #8704 opened 2 months ago by zyq213
  2

• Contribution Proposal: ScatterND Implementation for PyTorch

  #8703 opened 2 months ago by boubezariali
  5

• Time-consuming differences in different environments

  #8700 opened 2 months ago by v4if
  8

• torchvision is restricted to ffmpeg-4 on conda

  #8697 opened 3 months ago by bschindler
  5

• [Feature Request] PadToSquare: Square Padding to Preserve Aspect Ratios When Resizing Images with Varied Shapes in torchvision.transforms.v2

  #8699 opened 3 months ago by geezah
  4

• Pytorch build with rocm failed, execution_policy.h' file not found

  #8695 opened 3 months ago by FlintWangacc
  1

• torchvision.io.decode_image can not be used as document

  #8698 opened 3 months ago by malfet
  1

• PyTorch & Torchvision compatible issue on Jetson Orin

  #8696 opened 3 months ago by lida2003
  0

• CutMix/MixUp arguments

  #8693 opened 3 months ago by skrbnv
  2

• [RFC] Intel GPU (XPU) Support with Triton Implementation

  #8679 opened 3 months ago by Stonepia
  2

• Vision Transformer forward function

  #8685 opened 3 months ago by isega24
  3

• suppress `download_url` printing of "Using downloaded and verified file: ..."

  #8682 opened 3 months ago by talcs
  0

• Suppressing CIFAR uninformative and annoying download notification

  #8680 opened 3 months ago by talcs
  2

• Deterministic order of classes and images for the Omniglot dataset

  #8677 opened 3 months ago by V0XNIHILI
  0