ODinW: Results as Published

ODinW, Object Detection in the Wild, is the transfer setting in the GLIP paper. It uses 13 public object-detection datasets from Roboflow, each asking the model to localize a different set of objects, including drone views, aquarium scenes, ego-centric hands, and thermal images. The paper reports average precision on this collection. The same table includes a zero-shot column and several few-shot columns.

Every number here is read from the same catalog file as the benchmark table, data version 2026103001, and is the number the table shows for that model and test. Where one page printed a model under several setups, each value is listed. The catalog is public in armanobosyan/sugra-llm-benchmarks; a wrong number can be reported from the table, with Report an error under the value, or as an issue in that repository.