pytorch / pytorch/vision

Implementing and training SE-ResNet and SE-ResNeXt and including them in "torchvision.models"

Open
#2,179 3 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

module: models new feature
Dominant language
Python
Stars
17.9k
Forks
7.3k
Avg merge
1d 15h
Merged PRs (30d)
13

Description

🚀 Feature

Inclusion of pretrained SE-ResNet and SE-ResNeXt models in torchvision.

Motivation

SE-ResNet and SE-ResNeXt outperform ResNet and ResNeXt respectively and have been released around 2 years ago already.

Pitch

Implementing SE blocks and creating ResNet and ResNeXt models that include them, which will be called SE-ResNet and SE-ResNeXt and will be available as a part of torchvision.models. The SE paper has been published in CVPR 2018.

Alternatives

Additional context

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by reviewing the existing torchvision.models implementations and the SE paper linked in the issue to determine the required SE-ResNet and SE-ResNeXt variants. Done means the models are implemented, trained or provided with pretrained weights, and exposed through torchvision.models with appropriate validation.

Written by the indexing model from the issue text.

Assessment

Tech stack
python, pytorch
Domain
computer-vision, machine-learning
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.