awesome-grounding

Why we included this project

Visual grounding research spans a lot of subproblems, and this list does the sorting for you. It arranges papers into a chronological roadmap, with separate sections for referring expressions in images, natural language object retrieval, video activity localization, grounded description, pretraining, and 3D grounding, so you can see how the field developed instead of wading through a random pile of papers. Each entry links to the paper and, when available, the official code and dataset, which helps when you want to reproduce a baseline or compare methods. The maintainer says every paper was read before being added, and the contribution guidelines keep the curation honest. It also points to related multimodal and temporal grounding lists, making it a handy hub for scoping a project or building a literature review.