rf-zero Computer Vision Project

Roboflow

Updated a year ago

5

views

1

download
Classes (24)
Cell
Cell-Multi
No-Anomaly
Shadowing
Unclassified
avatar
barcode crack forklift gloves helmet ladder
paragraph
person pipe
red-blood-cell
safety vest
section
smoke table
traffic cone
wall-crack
white-blood-cell
wooden pallet
Description

Roboflow Zero (rf-zero) is an LMM evaluation dataset. You can use Roboflow Zero to evaluate LMMs across a range of different image types. This dataset features one image from the seven main categories in Roboflow 100:

  1. Aerial imagery
  2. Video games
  3. Microscopic data
  4. Underwater
  5. Documents
  6. Electromagnetic readings
  7. Common, real-world objects

We have also added images that relate to:

  1. Manufacturing
  2. Rail
  3. Security cameras
  4. Retail
  5. Pharmaceuticals (for use in pill counting)

You can use the images in Roboflow Zero with your own prompts to evaluate LMMs across a wide range of data. You can also use images in Roboflow Zero with the prompts associated with each image that we run for our LMM analyses. We publish LMM analyses on the Roboflow blog.

Supervision

Build Computer Vision Applications Faster with Supervision

Visualize and process your model results with our reusable computer vision tools.

Cite This Project

If you use this dataset in a research paper, please cite it using the following BibTeX:

                        @misc{
                            rf-zero_dataset,
                            title = { rf-zero Dataset },
                            type = { Open Source Dataset },
                            author = { Roboflow },
                            howpublished = { \url{ https://universe.roboflow.com/roboflow-gw7yv/rf-zero } },
                            url = { https://universe.roboflow.com/roboflow-gw7yv/rf-zero },
                            journal = { Roboflow Universe },
                            publisher = { Roboflow },
                            year = { 2023 },
                            month = { dec },
                            note = { visited on 2024-11-23 },
                            }
                        
                    

Similar Projects

See More