pytorch / pytorch/vision

Addition of OHEM sampler

Open
#4,286 10 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

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

Description

🚀 Feature

Addition of OHEM (online hard example mining) sampler

Motivation

Currently, a random sampler is being used in RoIHeads. However, other samplers have been shown to perform better than random sampling and heuristics. Thus, this motivates the need to introduce other samplers for example OHEM.

Pitch

There can be flag/option introduced to use a particular sampler. Based on the flag, the corresponding sampler would be used.

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 locating RoIHeads and the current random-sampler path. Read the surrounding model and sampling code to determine how a sampler option would fit, then review the issue discussion for expected behavior. Done means OHEM can be selected as an alternative sampler and its behavior is covered by appropriate tests.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
computer-vision, machine-learning
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.