Datasets:

Fraser
/

cursors

dataset_info:
  features:
    - name: cursor_type
      dtype: string
    - name: os
      dtype: string
    - name: frames
      list: image
    - name: frame_delay_ms
      dtype: int32
    - name: hotspot_x
      dtype: float32
    - name: hotspot_y
      dtype: float32
    - name: size_px
      dtype: int32
    - name: variant
      dtype: string
  splits:
    - name: train
      num_bytes: 9006656
      num_examples: 366
  download_size: 8991436
  dataset_size: 9006656
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
license: mit
tags:
  - cursors
  - gui
  - synthetic-data
  - grounding

Cursors

A dataset of labelled OS cursor images for compositing onto cursor-free screenshots as synthetic training data for GUI grounding models.

366 rows across 7 OS/theme sources, covering 31 cursor types.

Intended use

Given a screenshot with a cursor composited into it, a model is asked: "What cursor type is shown in the red box?"

The cursor_type labels are chosen to be visually distinguishable — if two themes render a cursor name differently (e.g. one theme draws move as a four-way arrow and another as a closed fist), the label reflects what the cursor actually looks like, not what the X11 name says.

Cursor type taxonomy

Pointers

cursor_type	Description	Typical appearance
`arrow`	Default pointer	Angled arrow pointing upper-left
`pointer`	Clickable / link	Hand with index finger pointing
`text`	Text selection	Vertical I-beam
`text_vertical`	Vertical text selection	Horizontal I-beam
`crosshair`	Precision select	Thin cross / plus

Drag & drop

cursor_type	Description	Typical appearance
`grab`	Element is grabbable	Open hand, palm facing out
`grabbing`	Actively dragging	Closed / clenched hand
`move`	Movable element	Four-way arrow cross
`copy`	Drag-copy	Arrow with small + badge
`alias`	Create shortcut / link	Arrow with curved-arrow badge

Resize

cursor_type	Description	Typical appearance
`resize_n`	North (up)	Upward arrow
`resize_s`	South (down)	Downward arrow
`resize_e`	East (right)	Rightward arrow
`resize_w`	West (left)	Leftward arrow
`resize_ne`	Northeast	Arrow pointing upper-right
`resize_nw`	Northwest	Arrow pointing upper-left
`resize_se`	Southeast	Arrow pointing lower-right
`resize_sw`	Southwest	Arrow pointing lower-left
`resize_ew`	Horizontal	Left-right double arrow
`resize_ns`	Vertical	Up-down double arrow
`resize_nesw`	Diagonal (/ axis)	Double arrow on NE-SW axis
`resize_nwse`	Diagonal (\ axis)	Double arrow on NW-SE axis

Status

cursor_type	Description	Typical appearance
`wait`	Busy / loading	Spinning circle or hourglass
`progress`	Busy but interactive	Arrow with spinner

Other

cursor_type	Description	Typical appearance
`not_allowed`	Forbidden / no-drop	Circle with diagonal line
`help`	Help available	Arrow with ? badge
`context_menu`	Right-click menu	Arrow with menu icon
`cell`	Cell / spreadsheet select	Hollow plus / crosshair
`zoom_in`	Zoom in	Magnifying glass with +
`zoom_out`	Zoom out	Magnifying glass with −
`all_scroll`	Omnidirectional scroll	Four-way arrow cluster

Schema

Field	Type	Description
`cursor_type`	string	One of the 31 types above
`os`	string	Source OS/theme identifier
`frames`	list[image]	Cursor image(s). Static cursors have 1 frame; animated cursors (wait, progress) have multiple
`frame_delay_ms`	int32	Milliseconds between animation frames (0 for static)
`hotspot_x`	float32	Click-point x position, normalised 0–1 within the image
`hotspot_y`	float32	Click-point y position, normalised 0–1 within the image
`size_px`	int32	Pixel dimensions of the cursor image (square)
`variant`	string	Colour variant: `default`, `black`, `white`, `blue`, `red`

Sources

`os` value	Source	Cursor sizes	Variants
`macos`	System cursors (PDF + NSCursor)	96px rendered	default
`linux-adwaita`	Xcursor binaries with hotspot metadata	96px	default
`linux-apple-cursor`	SVG-rendered Apple-style theme	96px	black, white
`linux-bibata`	SVG-rendered Bibata theme	96px	black, white
`linux-google-cursor`	Google Cursor pre-built PNGs	200px	black, white, blue, red
`linux-notwaita`	SVG-rendered Notwaita theme	96px	black, white
`linux-xcursor-pro`	SVG-rendered Xcursor-Pro theme	96px	black, white

Usage

from datasets import load_dataset

ds = load_dataset("Fraser/cursors", split="train")

# Get all arrow cursors
arrows = ds.filter(lambda r: r["cursor_type"] == "arrow")

# Get a specific cursor image
row = ds[0]
cursor_image = row["frames"][0]  # PIL Image
hotspot = (row["hotspot_x"], row["hotspot_y"])

Compositing a cursor onto a screenshot

Place a cursor so its hotspot (the click-point) lands at a target pixel position. The cursor is scaled to a realistic size and alpha-composited without artifacts.

from datasets import load_dataset
from PIL import Image


def composite_cursor(screenshot, cursor_frame, hotspot_x, hotspot_y,
                     target_x, target_y, cursor_size=32):
    """Place a cursor on a screenshot with the hotspot at (target_x, target_y)."""
    canvas = screenshot.convert("RGBA")
    cw, ch = cursor_frame.size
    scale = cursor_size / max(cw, ch)
    new_w, new_h = max(1, round(cw * scale)), max(1, round(ch * scale))
    cursor_scaled = cursor_frame.convert("RGBA").resize((new_w, new_h), Image.LANCZOS)

    # Offset so the hotspot lands on the target position
    paste_x = target_x - round(hotspot_x * new_w)
    paste_y = target_y - round(hotspot_y * new_h)

    # Crop if the cursor extends past the screenshot edges
    sw, sh = canvas.size
    src_left = max(0, -paste_x)
    src_top = max(0, -paste_y)
    src_right = min(new_w, sw - paste_x)
    src_bottom = min(new_h, sh - paste_y)
    if src_left >= src_right or src_top >= src_bottom:
        return screenshot
    cropped = cursor_scaled.crop((src_left, src_top, src_right, src_bottom))

    canvas.alpha_composite(cropped, (paste_x + src_left, paste_y + src_top))
    return canvas.convert(screenshot.mode)


ds = load_dataset("Fraser/cursors", split="train")

# Pick an arrow cursor from the Adwaita theme
row = ds.filter(lambda r: r["cursor_type"] == "arrow" and r["os"] == "linux-adwaita")[0]

screenshot = Image.open("my_screenshot.png")
result = composite_cursor(
    screenshot,
    row["frames"][0],
    row["hotspot_x"], row["hotspot_y"],
    target_x=500, target_y=300,  # where the arrow tip should point
    cursor_size=32,              # realistic size for a 1080p screenshot
)
result.save("my_screenshot_with_cursor.png")