lightly-ai / lightly-ai/lightly
CycleCL
Open
Nobody has claimed this yet.
feature request
- Dominant language
- Python
- Stars
- 3.8k
- Forks
- 367
- Avg merge
- 3d 22h
- Merged PRs (30d)
- 5
Description
CycleCL is a new contrastive learning method for cyclic videos: https://arxiv.org/abs/2311.03402
It could be an interesting addition to the package. We would have to add the following:
- [ ] Loading batches from a single video (ordered!)
- Finding negative and positive examples in a batch
- Loss -> It is basically Ntxentloss but generalized to accepting custom negative/positive pairs.
- Augmentations
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by reading the CycleCL paper linked in the issue and comparing its requirements with the package's existing NT-Xent loss and augmentation APIs. Define how ordered single-video batches, custom positive and negative pairs, the generalized loss, and augmentations should integrate. Done means all four checklist items are implemented and covered by appropriate tests.
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