A CVPRW survey turned awesome-list for text-to-image research

What it does This repository is a curated awesome-list that catalogs research papers, open-source projects, evaluation metrics, and datasets for text-to-image synthesis and manipulation. It functions as a living literature index accompanying a CVPRW 2023 survey on vision-and-language applications. The maintainer splits resources by year and sub-topic—ranging from general synthesis to niche areas like text-to-face generation.

The interesting bit Unlike most GitHub awesome-lists that accumulate links until they become unreadable, this one carries the credibility of a peer-reviewed survey paper and is currently undergoing a “Version 2.0” restructure to separate papers-with-code from other resources. That academic lineage gives it a slightly more rigorous editorial filter than your average star-hoarding list.

Key highlights

• Anchored by the CVPRW 2023 paper Vision+ Language Applications: A Survey.
• Organizes papers chronologically and by topic, including dedicated tracks for surveys, text-to-face, and “specific issues.”
• Maintains standalone lists for quantitative evaluation metrics and datasets.
• Highlights a “Best Collection” and a “Recently Focused Papers” shortlist for quick entry points.
• Version 2.0 (active since February 2024) splits content across separate markdown files to reduce noise.

Caveats

• Several sections remain unchecked on the maintainer’s todo list, including topic-order and chronological-order indexes.
• Paper counts for 2024 and 2025 currently show xx placeholders, so the latest literature is not yet tallied.

Verdict Grab this if you are a researcher or engineer trying to navigate the flood of text-to-image literature without drowning in arXiv noise. Skip it if you are hunting for a unified codebase or model weights—this is strictly a reading list and index.