UIPro-MultiUI-Data-v1

Part of the UIPro GUI-agent training suite (ICCV 2025). This repository packages the MultiUI source into the unified UIPro instruction-tuning format, with coordinates normalized to a [0, 1000] grid.

Web UI grounding, OCR, web QA and embedded QA derived from MultiUI.

Dataset at a glance

Total samples	3,410,901
Valid images	3,410,901
Avg. samples / image	1.0
Coordinate scale	0–1000
Source dataset	MultiUI

Samples by task

Task	Count
IntentGnd	1,328,726
EmbedQA	934,059
ElemGnd	729,642
OCR	207,062
TitleOCR	164,420
WebQA	46,992

Repository file structure

File	Description
MultiUI_gnd_ref_qa_scale1000_3410k.json	The dataset: a JSON list of 3,410,901 sample objects (schema below).
MultiUI_gnd_ref_qa_scale1000_3410k_sample.json	A small preview slice of the same schema, for quick inspection without downloading everything.
MultiUI_gnd_ref_qa_scale1000_3410k_images.zip	All screenshots referenced by the image field, preserving the relative paths stored there.
MultiUI_gnd_ref_qa_scale1000_3410k_info.json	Full generation report — per-task counts, image statistics, invalid-element breakdown, and the exact processing config.

Unzip MultiUI_gnd_ref_qa_scale1000_3410k_images.zip and each sample's image path resolves relative to the extraction root.

Sample schema — every field explained

Each element of the main JSON list is one training sample. This dataset's samples use the following fields:

Field	Meaning
conversations	The vision-language dialogue: a list of turns, each `{"from": "human"
id	Unique sample identifier, formatted autogui_<dataset>_<task>_<n>. The <task> segment (e.g. intentgnd, textloc, ocr, elemgnd, elemref) tells you which task the sample belongs to.
image	Path to the screenshot inside _images.zip, relative to the archive root. Load the image by joining this path with your extraction directory.
task_attr	The task's target attribute in plain form — for grounding tasks the referred element's text/instruction; for OCR/referring tasks the queried coordinate string. Useful for filtering or building custom prompts without parsing the conversation.
unnormalized_box	Ground-truth bounding box in original image pixels, as [x1, y1, x2, y2] (top-left, bottom-right). Present when a box is available. Note: the answer in the gpt turn is normalized to 0–1000, while this field is the raw-pixel box — divide by width/height and multiply by 1000 to reconcile them.

Note: Fields tied to a bounding box (e.g. unnormalized_box) are only present on samples that have a box; point-only answers (e.g. some intent-grounding samples) may omit them.

The conversations field in detail

conversations is a list of turns that a vision-language model consumes directly:

• Each turn is {"from": "...", "value": "..."}.
• from is either human (the prompt) or gpt (the ground-truth response).
• The token <image> inside a human turn marks where the screenshot is spliced into the prompt — replace it with the actual image when tokenizing.

Coordinate system

• Answers are normalized to the 0–1000 range relative to image width/height.
• A point answer is formatted (x,y); a bounding box answer is (x1,y1,x2,y2).
• Prompts ending in (with point) expect a point; (with bbox) expect a box.
• unnormalized_box, when present, is the same box in raw pixels — combine it with wxh (or the image's true size) to convert between pixels and the normalized grid.

Example

{
  "id": "autogui_multiui_embedqa_873493",
  "image": "MultiUI/v0.8_exclude_v0.7/caption_data_new/Training9_pc1280_curated_start_idx300000/data/002908721/0_embed.png",
  "conversations": [
    {
      "from": "human",
      "value": "<image>\nHow many eyes does the robot have?\nPlease answer the question with a detailed and comprehensive explanation."
    },
    {
      "from": "gpt",
      "value": "The caption describes the robot's head as having two circular eyes that convey a friendly expression, indicating that the robot has a pair of eyes."
    }
  ]
}

Usage

import json, os, zipfile
from PIL import Image
from huggingface_hub import hf_hub_download

repo = "HongxinLi/UIPro-MultiUI-Data-v1"
samples = json.load(open(hf_hub_download(repo, "MultiUI_gnd_ref_qa_scale1000_3410k.json", repo_type="dataset")))

images_zip = hf_hub_download(repo, "MultiUI_gnd_ref_qa_scale1000_3410k_images.zip", repo_type="dataset")
with zipfile.ZipFile(images_zip) as zf:
    zf.extractall("images/")

s = samples[0]
print(s["conversations"])
img = Image.open(os.path.join("images", s["image"]))   # screenshot for this sample
print("image size:", img.size)

About UIPro

UIPro is a generalist GUI agent trained on 20.6M understanding tasks across 13 task types, followed by agent continued fine-tuning. See the project repository for the full data pipeline, training recipes and evaluation scripts: https://github.com/ZJULiHongxin/UIPro

License

Released under CC BY-NC 4.0 (non-commercial research use). The underlying screenshots and annotations remain subject to the terms of their original source, MultiUI.

Citation

@inproceedings{uipro2025,
  title     = {UIPro: A Generalist GUI Agent},
  author    = {Li, Hongxin and others},
  booktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)},
  year      = {2025}
}