Vasilis Vryniotis

Results 15 issues of Vasilis Vryniotis

This is the continuation of the work from #6198 to finalize the API of MViT class. The PR extends TorchVision's existing MViT architecture to support [v2](https://arxiv.org/abs/2112.01526) variants. The `mvit_v2_s` variant...

enhancement
module: models
cla signed

### 🚀 The feature *Note: To track the progress of the project check out [this board](https://github.com/pytorch/vision/projects/6).* This is the 3rd phase of TorchVision's modernization project (see phase [1](https://github.com/pytorch/vision/issues/3911) and [2](https://github.com/pytorch/vision/issues/5410))....

help wanted
module: ops
module: models
module: transforms
module: reference scripts
new feature

### 🐛 Describe the bug ```python from torchvision.io import read_video from torchvision import set_video_backend path = "./3caPS4FHFF8_000036_000046.mp4" set_video_backend("pyav") video, audio, info = read_video(path) print("pyav OK") set_video_backend("video_reader") video, audio, info =...

bug
module: video
triage review
high priority

### 🚀 Feedback Request This issue is dedicated for collecting community feedback on the Model Registration API. Please review the dedicated [RFC](https://github.com/pytorch/vision/pull/6330) and [blogpost](https://pytorch.org/blog/easily-list-and-initialize-models-with-new-apis-in-torchvision/) where we describe the API in...

### 📚 Feedback Request This issue is dedicated for collecting community feedback on the Documentation revamping efforts. Here are some highlighted changes: - **New documentation layout:** Each function / class...

### 🚀 Feedback Request This issue is dedicated for collecting community feedback on the Multi-weight support API. Please review the [dedicated article](https://pytorch.org/blog/introducing-torchvision-new-multi-weight-support-api/) where we describe the API in detail and...

### 🚀 The feature Swin Transforms were added to TorchVision on #5491 thanks to the work of @xiaohu2015 and @jdsgomes. At that point, the official code of [SwinV2](https://arxiv.org/abs/2111.09883) was not...

module: models
new feature

### 🚀 The feature Consider adding the following operators in TorchVision: ## Layers There is a separate ticket for tracking common layers: #4333 - [x] [DropBlock Layer](https://arxiv.org/abs/1810.12890) #5416 - [...

needs discussion
module: ops

### 🚀 The feature As discussed at https://github.com/pytorch/vision/issues/6517, passing `None` on `fill` for some transforms isn't the same as passing `0`. The difference applies on some cases but not others,...

needs discussion
module: transforms
bc-breaking

### 🐛 Describe the bug The `cmake_windows_cpu` job started [failing](https://app.circleci.com/pipelines/github/pytorch/vision/20092/workflows/3667f674-cf71-4f04-ab6a-044000cacbaa/jobs/1632954) on 20220912 with: ``` Specifications: - pytorch=1.13.0.dev20220912 -> python[version='>=3.10,=3.9,=3.8,=2021.4.0,=2018'] -> intel-openmp[version='2021.*|2022.*'] pytorch=1.13.0.dev20220912 -> intel-openmp Package ca-certificates conflicts for: python=3.7 ->...

dependency issue
high priority
module: ci
core issue