Upload Pointerbench v1 dataset

README.md

@@ -0,0 +1,110 @@
+---
+license: cc-by-4.0
+task_categories:
+- image-to-text
+- visual-question-answering
+language:
+- en
+- de
+- fr
+- es
+- it
+- nl
+tags:
+- gui-grounding
+- computer-use
+- benchmark
+- screenshots
+- synthetic-data
+- spreadsheets
+- text-grounding
+- professional-apps
+pretty_name: Pointerbench
+size_categories:
+- 1K<n<10K
+---
+
+# Pointerbench
+
+Pointerbench is a small GUI grounding benchmark suite for computer-use models.
+Each example has one screenshot, one instruction, target geometry in absolute
+pixels, and a binary evaluation rule.
+
+The suite has three subsets:
+
+| Subset | Examples | What it tests |
+| --- | ---: | --- |
+| `pointerbench-sheets` | 500 | Spreadsheet cells, colors, headers, edges, corners, and relative positions |
+| `pointerbench-text` | 500 | Words, characters, punctuation, caret positions, chrome text, and text bounding boxes |
+| `pointerbench-pro` | 500 | Icons, text, and mixed GUI targets across 100 professional applications |
+
+All images are synthetic 1024x768 PNG screenshots. The datasets contain no
+scraped user data and no PII.
+
+## Layout
+
+Each subset is self-contained:
+
+```text
+pointerbench-sheets/
+  data/test/metadata.jsonl
+  data/test/0000.png
+  eval.py
+  README.md
+  REPRODUCE.md
+
+pointerbench-text/
+  data/test/metadata.jsonl
+  data/test/0000.png
+  eval.py
+  README.md
+  REPRODUCE.md
+
+pointerbench-pro/
+  data/test/metadata.jsonl
+  data/test/0000.png
+  eval.py
+  README.md
+  REPRODUCE.md
+```
+
+## Schema
+
+Each metadata row includes:
+
+```json
+{
+  "file_name": "0000.png",
+  "id": "pbs_0000",
+  "instruction": "Click cell E11.",
+  "bbox": [x1, y1, x2, y2],
+  "point": [x, y],
+  "answer_type": "point",
+  "eval": {"type": "point_in_bbox", "bbox": [x1, y1, x2, y2]},
+  "data_type": "cell",
+  "category": "cell_ref",
+  "image_size": [1024, 768]
+}
+```
+
+Point tasks are correct when the predicted point lands inside the target bbox.
+Bbox tasks, used in Pointerbench-Text, are correct when the predicted bbox
+reaches the configured IoU threshold.
+
+## Evaluation
+
+Run the scorer inside a subset folder:
+
+```bash
+python eval.py --predictions preds.jsonl
+```
+
+Predictions are JSONL rows with an `id` and either a `point` or `bbox`, depending
+on `answer_type`.
+
+See each subset README for the exact distribution, schema details, and examples.
+
+## License
+
+Dataset images and annotations are released under CC BY 4.0. The included
+evaluation scripts are released under MIT.

pointerbench-pro/.gitignore

@@ -0,0 +1,8 @@
+__pycache__/
+*.pyc
+*.egg-info/
+.venv/
+.DS_Store
+# local prediction / report artifacts
+preds*.jsonl
+report*.json

pointerbench-pro/CITATION.cff

@@ -0,0 +1,9 @@
+cff-version: 1.2.0
+message: "If you use Pointerbench-Pro, please cite it as below."
+title: "Pointerbench-Pro: A GUI Grounding Benchmark for Professional Software"
+authors:
+  - name: "Pointerbench-Pro contributors"
+version: "1.0.0"
+date-released: "2026-06-18"
+url: "https://github.com/YOUR_ORG/pointerbench-pro"
+license: "CC-BY-4.0"

pointerbench-pro/LICENSE

@@ -0,0 +1,64 @@
+Attribution 4.0 International (CC BY 4.0)
+
+The dataset (images and annotations) in this repository is licensed under the
+Creative Commons Attribution 4.0 International License. The accompanying code
+(eval.py) is licensed under the MIT License (see below).
+
+================================================================================
+Creative Commons Attribution 4.0 International (CC BY 4.0)
+================================================================================
+
+You are free to:
+
+  Share: copy and redistribute the material in any medium or format.
+  Adapt: remix, transform, and build upon the material for any purpose, even
+         commercially.
+
+The licensor cannot revoke these freedoms as long as you follow the license
+terms.
+
+Under the following terms:
+
+  Attribution: You must give appropriate credit, provide a link to the
+  license, and indicate if changes were made. You may do so in any reasonable
+  manner, but not in any way that suggests the licensor endorses you or your
+  use.
+
+  No additional restrictions: You may not apply legal terms or technological
+  measures that legally restrict others from doing anything the license
+  permits.
+
+Notices:
+
+  You do not have to comply with the license for elements of the material in
+  the public domain or where your use is permitted by an applicable exception
+  or limitation.
+
+  No warranties are given. The license may not give you all of the permissions
+  necessary for your intended use. For example, other rights such as publicity,
+  privacy, or moral rights may limit how you use the material.
+
+The full legal text of CC BY 4.0 is available at:
+https://creativecommons.org/licenses/by/4.0/legalcode
+
+================================================================================
+MIT License (applies to code: eval.py)
+================================================================================
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

pointerbench-pro/README.md

@@ -0,0 +1,196 @@
+# Pointerbench-Pro
+
+**A 500-example GUI grounding benchmark for professional desktop software.**
+Given a screenshot of a professional application and a short functional
+instruction (e.g. *"Begin a performance recording."*, *"Select the Bass
+track."*, *"Open the Browser panel."*), a model must output the pixel coordinate
+to click. Scored exactly like
+[ScreenSpot](https://github.com/njucckevin/SeeClick) and ScreenSpot-Pro: **a
+click is correct if it lands inside the target's bounding box.**
+
+![teaser](assets/teaser.png)
+
+## Why professional apps?
+
+Everyday GUI suites are dominated by browsers and consumer apps. Real agent work
+happens in dense expert tools: IDEs, DAWs, CAD, GIS, EDA, data and analytics
+consoles, scientific software. Their interfaces are packed with small icons and
+tightly labeled controls, which is exactly where grounding breaks down.
+Pointerbench-Pro targets that regime across **100 professional applications**,
+with the generated target mix preserved instead of forcing a fixed icon/text
+quota.
+
+## What's inside
+
+- **500** examples, one instruction per image.
+- **1024x768** PNG screenshots, fully synthetic (model-dreamed UIs, no scraping,
+  no PII).
+- **100 applications** spanning development, creative, scientific,
+  CAD/engineering, productivity, enterprise, office, communication, analytics,
+  finance, and OS/system surfaces.
+- **3 target types from the generated source pools**:
+
+  | element_type | count | example instruction              |
+  | ------------ | ----- | -------------------------------- |
+  | `icon`       | 190   | *Find a packet.*                 |
+  | `other`      | 156   | *Select the red-coat clip.*      |
+  | `text`       | 154   | *Switch to Home.*                |
+
+- Multiple platforms (Windows, macOS, Linux, plus a few embedded/mobile).
+- Functional intent prompts (not literal element names), matching the
+  ScreenSpot-Pro task style.
+
+## Schema
+
+Each line of `data/test/metadata.jsonl` (HuggingFace `imagefolder` layout):
+
+```json
+{
+  "file_name": "0000.png",
+  "id": "pbp_0000",
+  "instruction": "Begin a performance recording.",
+  "bbox": [596, 376, 681, 395],
+  "point": [638, 385],
+  "answer_type": "point",
+  "eval": {"type": "point_in_bbox", "bbox": [596, 376, 681, 395]},
+  "data_type": "icon",
+  "element_type": "icon",
+  "app": "Google Chrome DevTools",
+  "app_slug": "chrome_devtools",
+  "app_category": "development",
+  "platform": "Windows 11 / Chrome",
+  "source_id": "images_chrome_devtools_02_el8",
+  "source_file": "clicks_icons_pro_3k.ndjson",
+  "image_size": [1024, 768]
+}
+```
+
+- **`bbox`**: ground-truth target, **`[x1, y1, x2, y2]` in absolute pixels**
+  (top-left, bottom-right) on the 1024x768 image. A prediction is correct iff it
+  lands inside this box.
+- **`point`**: the box center.
+- **`answer_type`** / **`eval`**: binary evaluation rule. Pro rows are point
+  answers scored with point-in-bbox.
+- **`element_type`** (= `data_type`): `icon`, `text`, or `other`.
+- **`app`** / **`app_slug`** / **`app_category`** / **`platform`**: the source
+  application and its metadata, for per-app result breakdowns.
+
+A machine-readable coverage summary (counts per app, category, platform) is in
+[`apps.json`](apps.json).
+
+## Quickstart
+
+### Load the data
+
+Via HuggingFace `datasets` (after the set is pushed to the Hub):
+
+```python
+from datasets import load_dataset
+ds = load_dataset("YOUR_ORG/pointerbench-pro", split="test")
+ex = ds[0]
+ex["image"]        # PIL.Image, 1024x768
+ex["instruction"]  # "Begin a performance recording."
+ex["bbox"]         # [x1, y1, x2, y2]
+```
+
+Or locally with the imagefolder loader:
+
+```python
+from datasets import load_dataset
+ds = load_dataset("imagefolder", data_dir="data", split="test")
+```
+
+Or with no dependencies at all, read `data/test/metadata.jsonl` and open the
+sibling PNGs yourself.
+
+### Evaluate
+
+1. Run your model on every example's `instruction` + image and collect a
+   predicted click point (absolute pixels on the 1024x768 image).
+2. Write predictions as JSONL, one object per example:
+
+   ```json
+   {"id": "pbp_0000", "point": [612, 388]}
+   ```
+
+3. Score (pure standard library, no dependencies):
+
+   ```bash
+   python eval.py --predictions preds.jsonl --json report.json
+   ```
+
+   ```
+   Pointerbench-Pro: 500 examples
+   ============================================
+   Accuracy: 41.20%   (206/500)
+
+   By target type:
+    icon                     33.20%   (n=190)
+    other                    37.18%   (n=156)
+    text                     49.35%   (n=154)
+     ...
+   ```
+
+The scorer reports overall accuracy plus per-target-type, per-app-category, and
+per-platform breakdowns; the per-app table is in the `--json` report.
+
+### Turning model output into a point
+
+Models emit clicks in many shapes; map them to `[x, y]` pixels before scoring.
+For example, a `<click>x,y</click>` tag or a normalized `0-1` / `0-999` point:
+
+```python
+import re
+def to_point(text, w=1024, h=768):
+    m = re.search(r"(-?\d+(?:\.\d+)?)\s*[,\s]\s*(-?\d+(?:\.\d+)?)", text)
+    x, y = float(m.group(1)), float(m.group(2))
+    if max(x, y) <= 1.0:   x, y = x * w, y * h          # normalized 0-1
+    elif max(x, y) <= 999: x, y = x / 999 * w, y / 999 * h  # 0-999 grid
+    return [round(x), round(y)]
+```
+
+## Baselines
+
+| Model                         | Accuracy | Notes                     |
+| ----------------------------- | -------- | ------------------------- |
+| Center-click (512, 384)       | low      | sanity floor              |
+| _your model here_             | n/a      | open a PR                 |
+
+## Construction
+
+- **Screenshots** are dreamed by an image model from per-app scenario prompts
+  (one scene per professional application), so the UIs are realistic but
+  contain no real user data.
+- **Boxes** are pixel-exact: a copy of each screenshot is edited to paint the
+  target region, and the box is recovered by diffing the edit against the clean
+  image. QA gates reject ambiguous or oversized regions.
+- Targets come from the labeled professional-icon pool and the general
+  professional GUI intent pool. They are labeled as `icon`, `text`, or `other`
+  for analysis, but they are not filtered to a fixed element-type ratio.
+
+The selected `source_id`s are listed in `heldout_source_ids.txt` so they can be
+excluded from any training set built from the same generator. See
+[`REPRODUCE.md`](REPRODUCE.md).
+
+## Limitations
+
+- Fully synthetic: realistic but not screenshots of the real applications.
+- Fixed 1024x768 resolution; instructions in English.
+- The `text` / `other` label for the general intent pool is derived by a
+  description heuristic rather than human annotation.
+
+## Citation
+
+```bibtex
+@misc{pointerbench_pro_2026,
+  title  = {Pointerbench-Pro: A GUI Grounding Benchmark for Professional Software},
+  author = {Pointerbench-Pro contributors},
+  year   = {2026},
+  url    = {https://github.com/YOUR_ORG/pointerbench-pro}
+}
+```
+
+## License
+
+- **Data** (images + annotations): [CC BY 4.0](LICENSE).
+- **Code** (`eval.py`): MIT.

pointerbench-pro/REPRODUCE.md

@@ -0,0 +1,31 @@
+# Reproducing Pointerbench-Pro
+
+The benchmark is assembled from the `dream_click` generator's professional-app
+click pools in the source repository.
+
+```bash
+# from the generator directory in the source repo:
+python3 build_progui.py --n 500 --seed 90311 --out /path/to/pointerbench-pro
+```
+
+- Targets are sampled from the labeled professional-icon pool
+  (`clicks_icons_pro_3k.ndjson`) and the general dreamed-GUI intent pool
+  (`clicks_dreamed_gui_intent.ndjson`).
+- Targets are labeled as `icon`, `text`, or `other` for analysis, but the public
+  set does not force a fixed icon/text ratio.
+- Candidates are balanced across applications by round-robin selection with a
+  fixed seed, then shuffled. The current 500-example set contains 190 icon, 154
+  text, and 156 other targets across 100 applications.
+- Each application's name, slug, category, and platform come from the dreamed-GUI
+  scenario indexes under `shared/data/dreamed_gui*/index.jsonl`, joined by the
+  scene slug parsed from each `source_id`.
+
+## Held-out usage
+
+The examples are selected from existing generated pools, so they can overlap a
+training set built from the same generator. The selected `source_id`s are
+written to `heldout_source_ids.txt`. Exclude those IDs from training to keep the
+benchmark held out.
+
+Regenerating with the same seed reproduces the identical 500 examples. To grow
+or refresh the set, change `--n` / `--seed` and bump the dataset version.

pointerbench-pro/apps.json

@@ -0,0 +1,150 @@
+{
+  "n": 500,
+  "element_type": {
+    "other": 156,
+    "text": 154,
+    "icon": 190
+  },
+  "n_apps": 100,
+  "platform": {
+    "Windows 11": 190,
+    "macOS Sonoma": 65,
+    "Windows 10": 55,
+    "Windows 11 / Chrome": 40,
+    "Windows 11, Chrome": 35,
+    "macOS": 30,
+    "Linux": 20,
+    "macOS / Chrome": 15,
+    "Linux / Firefox": 10,
+    "macOS / Safari": 5,
+    "Embedded automotive": 5,
+    "Ubuntu 24.04": 5,
+    "Embedded touchscreen": 5,
+    "iPadOS": 5,
+    "Linux / Chrome": 5,
+    "macOS Sonoma (desktop app)": 5,
+    "Windows 11 / Edge": 5
+  },
+  "app_category": {
+    "development": 105,
+    "creative": 95,
+    "cad_engineering": 45,
+    "productivity": 45,
+    "scientific": 45,
+    "consumer": 40,
+    "enterprise": 30,
+    "office": 25,
+    "communication": 25,
+    "os_commons": 20,
+    "analytics": 10,
+    "finance": 5,
+    "smart_home": 5,
+    "cms": 5
+  },
+  "source_file": {
+    "clicks_dreamed_gui_intent.ndjson": 310,
+    "clicks_icons_pro_3k.ndjson": 190
+  },
+  "apps": {
+    "Adobe Premiere Pro": 5,
+    "Microsoft Excel (Office 365)": 5,
+    "desktop stock trading platform (thinkorswim-style)": 5,
+    "Canva": 5,
+    "Mozilla Thunderbird": 5,
+    "Wireshark": 5,
+    "Discord desktop": 5,
+    "Unreal Engine 5": 5,
+    "DBeaver": 5,
+    "KiCad 8 PCB Editor": 5,
+    "Asana": 5,
+    "SkyLink Airways (fictional)": 5,
+    "Tableau Desktop": 5,
+    "VMware Fusion": 5,
+    "DriveOS Infotainment (fictional)": 5,
+    "Godot Engine": 5,
+    "RStudio": 5,
+    "Autodesk Fusion 360": 5,
+    "AutoCAD Mechanical": 5,
+    "Adobe Illustrator": 5,
+    "SolidWorks Premium": 5,
+    "MongoDB Compass": 5,
+    "Spotify desktop app": 5,
+    "Zoom desktop": 5,
+    "Stata SE 16": 5,
+    "Notion desktop": 5,
+    "Microsoft PowerPoint 2019": 5,
+    "Netflix (web)": 5,
+    "Microsoft Teams": 5,
+    "macOS Sonoma system UI": 5,
+    "REAPER": 5,
+    "Google Analytics 4 in Chrome": 5,
+    "DaVinci Resolve 19": 5,
+    "IntelliJ IDEA": 5,
+    "Unity Editor": 5,
+    "Airtable": 5,
+    "Audacity": 5,
+    "Inkscape": 5,
+    "GIMP": 5,
+    "MediChart EHR (fictional)": 5,
+    "Autodesk Revit": 5,
+    "SketchUp Pro": 5,
+    "YouTube Studio": 5,
+    "Adobe Lightroom Classic": 5,
+    "Microsoft Word (Office 365)": 5,
+    "Krita": 5,
+    "Android Studio": 5,
+    "Xcode": 5,
+    "MATLAB R2022b": 5,
+    "Adobe Photoshop": 5,
+    "Trello": 5,
+    "Home Assistant dashboard in Chrome": 5,
+    "cPanel": 5,
+    "Jira Software in Chrome": 5,
+    "Windows 11 system UI": 5,
+    "AWS Management Console": 5,
+    "CapCut Desktop": 5,
+    "Slack desktop": 5,
+    "Salesforce Lightning (CRM) in Chrome": 5,
+    "SAP GUI (ERP)": 5,
+    "Ubuntu 24.04 GNOME desktop": 5,
+    "Microsoft Outlook": 5,
+    "OrderHub POS (fictional)": 5,
+    "PyCharm": 5,
+    "Postman": 5,
+    "chess.com in a browser": 5,
+    "Intel Quartus II": 5,
+    "iOS Settings (iPad)": 5,
+    "Obsidian": 5,
+    "Proxmox VE": 5,
+    "GitHub (web)": 5,
+    "Steam desktop client": 5,
+    "QGIS": 5,
+    "Miro": 5,
+    "VLC Media Player": 5,
+    "OriginLab Origin 2018": 5,
+    "online banking web portal in a browser": 5,
+    "Linear": 5,
+    "Grafana": 5,
+    "Figma": 5,
+    "Adobe Acrobat Pro": 5,
+    "Xilinx Vivado": 5,
+    "Anki": 5,
+    "FL Studio 20": 5,
+    "Microsoft Azure Portal": 5,
+    "Ableton Live 12": 5,
+    "JupyterLab": 5,
+    "Sublime Text": 5,
+    "Blender 4": 5,
+    "IBM SPSS Statistics": 5,
+    "Docker Desktop": 5,
+    "Autodesk Inventor Professional": 5,
+    "WordPress admin in Chrome": 5,
+    "EViews 10": 5,
+    "NI LabVIEW": 5,
+    "Lens (Kubernetes IDE)": 5,
+    "Visual Studio Code": 5,
+    "UltiMaker Cura 5": 5,
+    "Google Chrome DevTools": 5,
+    "OBS Studio": 5
+  }
+}