Datasets:
metadata
annotations_creators:
- no-annotation
language:
- en
language_creators:
- found
license:
- other
multilinguality:
- monolingual
pretty_name: crello
size_categories:
- 10K<n<100K
source_datasets:
- original
tags:
- graphic design
- design templates
task_categories:
- unconditional-image-generation
task_ids: []
dataset_info:
features:
- name: id
dtype: string
- name: length
dtype: int64
- name: group
dtype:
class_label:
names:
'0': BG
'1': EO
'2': HC
'3': MM
'4': SM
'5': SMA
- name: format
dtype:
class_label:
names:
'0': Album Cover
'1': Book Cover
'2': Brochure
'3': Business card
'4': Calendar
'5': Card
'6': Certificate
'7': Coupon
'8': Email header
'9': FB event cover
'10': Facebook
'11': Facebook AD
'12': Facebook cover
'13': Flayer
'14': Gallery Image
'15': Gift Certificate
'16': Graphic
'17': IGTV Cover
'18': Image
'19': Infographic
'20': Instagram
'21': Instagram AD
'22': Instagram Highlight Cover
'23': Instagram Story
'24': Invitation
'25': Invoice
'26': Label
'27': Large Rectangle
'28': Leaderboard
'29': Letterhead
'30': LinkedIn Cover
'31': Logo
'32': Medium Rectangle
'33': Menu
'34': Mind Map
'35': Mobile Presentation
'36': Mood Board
'37': Newsletter
'38': Photo Book
'39': Pinterest
'40': Postcard
'41': Poster
'42': Poster US
'43': Presentation
'44': Presentation Wide
'45': Proposal
'46': Recipe Card
'47': Resume
'48': Schedule Planner
'49': Skyscraper
'50': Snapchat Geofilter
'51': Snapchat Moment Filter
'52': Storyboard
'53': T-Shirt
'54': Ticket
'55': Title
'56': Tumblr
'57': Twitch Offline Banner
'58': Twitch Profile Banner
'59': Twitter
'60': VK Community Cover
'61': VK Post with Button
'62': VK Universal Post
'63': Web Banner
'64': Youtube
'65': Youtube Thumbnail
'66': Zoom Background
- name: canvas_width
dtype:
class_label:
names:
'0': '1000'
'1': '1008'
'2': '1024'
'3': '1080'
'4': '1128'
'5': '1190'
'6': '1200'
'7': '1280'
'8': '1296'
'9': '1500'
'10': '1590'
'11': '160'
'12': '1600'
'13': '1920'
'14': '240'
'15': '241'
'16': '2560'
'17': '300'
'18': '3000'
'19': '336'
'20': '360'
'21': '396'
'22': '419'
'23': '420'
'24': '432'
'25': '500'
'26': '537'
'27': '540'
'28': '560'
'29': '576'
'30': '595'
'31': '600'
'32': '635'
'33': '728'
'34': '735'
'35': '792'
'36': '800'
'37': '841'
'38': '842'
'39': '851'
'40': '940'
- name: canvas_height
dtype:
class_label:
names:
'0': '1055'
'1': '1080'
'2': '1102'
'3': '1200'
'4': '1296'
'5': '141'
'6': '142'
'7': '1440'
'8': '1600'
'9': '1683'
'10': '1728'
'11': '191'
'12': '1920'
'13': '200'
'14': '2000'
'15': '216'
'16': '2340'
'17': '240'
'18': '250'
'19': '2560'
'20': '280'
'21': '288'
'22': '297'
'23': '298'
'24': '315'
'25': '320'
'26': '380'
'27': '400'
'28': '480'
'29': '500'
'30': '504'
'31': '512'
'32': '576'
'33': '595'
'34': '600'
'35': '612'
'36': '628'
'37': '654'
'38': '700'
'39': '720'
'40': '768'
'41': '788'
'42': '810'
'43': '841'
'44': '842'
'45': '90'
- name: category
dtype:
class_label:
names:
'0': all
'1': beauty
'2': businessFinance
'3': citiesPlaces
'4': educationScience
'5': fashionStyle
'6': foodDrinks
'7': handcraftArt
'8': holidaysCelebration
'9': homeStuff
'10': industry
'11': kidsParents
'12': leisureEntertainment
'13': medical
'14': natureWildlife
'15': pets
'16': realEstateBuilding
'17': religions
'18': socialActivityCharity
'19': sportExtreme
'20': technology
'21': transportation
'22': travelsVacations
- name: title
dtype: string
- name: type
sequence:
class_label:
names:
'0': coloredBackground
'1': imageElement
'2': maskElement
'3': svgElement
'4': textElement
- name: left
sequence: float32
- name: top
sequence: float32
- name: width
sequence: float32
- name: height
sequence: float32
- name: opacity
sequence: float32
- name: text
sequence: string
- name: font
sequence:
class_label:
names:
'0': ''
'1': Abril Fatface
'2': Aldrich
'3': Alef
'4': Alegreya Sans
'5': Alfa Slab One
'6': Alice
'7': Allerta Stencil
'8': Allura
'9': Amatic Sc
'10': Anton
'11': Arapey
'12': Architects Daughter
'13': Arima Madurai
'14': Arimo
'15': Arizonia
'16': Arkana Script
'17': Armata
'18': Assistant
'19': Bad Script
'20': Baloo Tamma
'21': Bangers
'22': Barrio
'23': Beacon
'24': Bebas Neue
'25': Bellefair
'26': Bentham
'27': Berkshire Swash
'28': Bilbo
'29': Black Ops One
'30': Blogger
'31': Breathe
'32': Breathe Press
'33': Brusher
'34': Brusher Free Font
'35': Bubbler One
'36': Buda
'37': Bungee
'38': Bungee Shade
'39': Cabin Sketch
'40': Caesar Dressing
'41': Cantarell
'42': Carter One
'43': Caveat
'44': Cedarville Cursive
'45': Chathura
'46': Clicker Script
'47': Comfortaa
'48': Contrail One
'49': Cookie
'50': Copse
'51': Cormorant Infant
'52': Courgette
'53': Cousine
'54': Covered By Your Grace
'55': Crete Round
'56': Cutive Mono
'57': Damion
'58': Dancing Script
'59': David Libre
'60': Dawning Of A New Day
'61': Delius
'62': Delius Swash Caps
'63': Didact Gothic
'64': Dorsa
'65': Dosis
'66': Droid Serif
'67': Dukomdesign Constantine
'68': Eb Garamond
'69': Economica
'70': El Messiri
'71': Elsie
'72': Elsie Swash Caps
'73': Euphoria Script
'74': Ewert
'75': Exo 2
'76': Farsan
'77': Faster One
'78': Fauna One
'79': Finger Paint
'80': Fjalla One
'81': Forum
'82': Frank Ruhl Libre
'83': Fredericka The Great
'84': Gabriela
'85': Gaegu
'86': Geo
'87': Gfs Didot
'88': Give You Glory
'89': Glass Antiqua
'90': Gluk Glametrix
'91': Gluk Znikomitno25
'92': Graduate
'93': Grand Hotel
'94': Gravitas One
'95': Great Vibes
'96': Gruppo
'97': Handlee
'98': Happy Monkey
'99': Heebo
'100': Homemade Apple
'101': Iceberg
'102': Iceland
'103': Im Fell
'104': Im Fell Dw Pica Sc
'105': Inconsolata
'106': Italiana
'107': Italianno
'108': Jacques Francois Shadow
'109': Josefin Sans
'110': Josefin Slab
'111': Julius Sans One
'112': Junge
'113': Jura
'114': Just Me Again Down Here
'115': Kalam
'116': Katibeh
'117': Kaushan Script
'118': Kavivanar
'119': Kelly Slab
'120': Knewave
'121': Knewave Outline
'122': Kreon
'123': Kristi
'124': Kumar One
'125': Kumar One Outline
'126': Kurale
'127': La Belle Aurore
'128': Lalezar
'129': Lato
'130': Lauren
'131': League Script
'132': Lemon Tuesday
'133': Libre Baskerville
'134': Limelight
'135': Londrina Shadow
'136': Londrina Sketch
'137': Loved By The King
'138': Lovers Quarrel
'139': Marcellus Sc
'140': Marck Script
'141': Mate
'142': Maven Pro
'143': Meddon
'144': Medula One
'145': Merienda One
'146': Merriweather
'147': Mikodacs
'148': Miriam Libre
'149': Monda
'150': Monofett
'151': Monsieur La Doulaise
'152': Montserrat
'153': Montserrat Alternates
'154': Mr Dafoe
'155': Mr De Haviland
'156': Mrs Saint Delafield
'157': Mrs Sheppards
'158': Neucha
'159': Nixie One
'160': Nothing You Could Do
'161': Noticia Text
'162': Nova Square
'163': Nunito
'164': Offside
'165': Okolaks
'166': Old Standard Tt
'167': Oleo Script
'168': Open Sans
'169': Open Sans Condensed
'170': Oranienbaum
'171': Orbitron
'172': Oswald
'173': Overlock
'174': Oxygen
'175': Pacifico
'176': Pangolin
'177': Parisienne
'178': Pathway Gothic One
'179': Patrick Hand
'180': Pattaya
'181': Patua One
'182': Permanent Marker
'183': Petit Formal Script
'184': Philosopher
'185': Pinyon Script
'186': Pirou
'187': Play
'188': Playball
'189': Playfair Display
'190': Playlist Caps
'191': Playlist Script
'192': Podkova
'193': Poiret One
'194': Pompiere
'195': Port Lligat Slab
'196': Press Start 2P
'197': Prompt
'198': Pt Sans
'199': Quattrocento
'200': Quicksand
'201': Racing Sans One
'202': Radley
'203': Rakkas
'204': Raleway
'205': Raleway Dots
'206': Rammetto One
'207': Rationale
'208': Reem Kufi
'209': Reenie Beanie
'210': Righteous
'211': Rise
'212': Rissa Typeface
'213': Roboto
'214': Rochester
'215': Rock Salt
'216': Rokkitt
'217': Rosario
'218': Rubik
'219': Rubik One
'220': Ruslan Display
'221': Russo One
'222': Rye
'223': Sacramento
'224': Sansita One
'225': Satisfy
'226': Scope One
'227': Secular One
'228': Selima Script
'229': Sensei
'230': Seymour One
'231': Shadows Into Light Two
'232': Share Tech Mono
'233': Sirin Stencil
'234': Six Caps
'235': Source Serif Pro
'236': Space Mono
'237': Stalemate
'238': Stint Ultra Expanded
'239': Sue Ellen Francisco
'240': Suez One
'241': Sunday
'242': Superclarendon Regular
'243': Text Me One
'244': Tinos
'245': Titillium Web
'246': Tulpen One
'247': Underdog
'248': V T323
'249': Vampiro One
'250': Varela Round
'251': Vast Shadow
'252': Vollkorn
'253': Waiting For The Sunrise
'254': Wire One
'255': Yanone Kaffeesatz
'256': Yellowtail
'257': Yeseva One
'258': Yesteryear
'259': Zeyada
'260': Znikomit
'261': Znikomitno24
- name: font_size
sequence: float32
- name: text_align
sequence:
class_label:
names:
'0': ''
'1': center
'2': left
'3': right
- name: angle
sequence: float32
- name: capitalize
sequence:
class_label:
names:
'0': 'false'
'1': 'true'
- name: line_height
sequence: float32
- name: letter_spacing
sequence: float32
- name: suitability
sequence:
class_label:
names:
'0': mobile
- name: keywords
sequence: string
- name: industries
sequence:
class_label:
names:
'0': artCrafts
'1': beautyCosmetics
'2': businessFinance
'3': corporate
'4': ecologyNature
'5': educationTraining
'6': entertainmentLeisure
'7': familyKids
'8': fashionStyle
'9': foodBeverages
'10': healthWellness
'11': homeLiving
'12': hrRecruitment
'13': marketingAds
'14': nonProfitCharity
'15': petsAnimals
'16': realEstateConstruction
'17': religionFaith
'18': retail
'19': services
'20': sportFitness
'21': techGadgets
'22': transportDelivery
'23': travelTourism
- name: color
sequence:
sequence: float32
length: 3
- name: image
sequence: image
splits:
- name: train
num_bytes: 3322744283.141
num_examples: 18659
- name: test
num_bytes: 421990602.771
num_examples: 2371
- name: validation
num_bytes: 425905823.995
num_examples: 2391
download_size: 4130251706
dataset_size: 4170640709.9069996
Dataset Card for Crello
Table of Contents
- Dataset Card for Crello
Dataset Description
- Homepage: CanvasVAE github
- Repository:
- Paper: CanvasVAE: Learning to Generate Vector Graphic Documents
- Leaderboard:
- Point of Contact: Kota Yamaguchi
Dataset Summary
The Crello dataset is compiled for the study of vector graphic documents. The dataset contains document meta-data such as canvas size and pre-rendered elements such as images or text boxes. The original templates were collected from crello.com (now create.vista.com) and converted to a low-resolution format suitable for machine learning analysis.
Supported Tasks and Leaderboards
CanvasVAE studies unsupervised document generation.
Languages
Almost all design templates use English.
Dataset Structure
Data Instances
Each instance has scalar attributes (canvas) and sequence attributes (elements). Categorical values are stored as integer values. Check ClassLabel features of the dataset for the list of categorical labels.
{'id': '592d6c2c95a7a863ddcda140',
'length': 8,
'group': 4,
'format': 20,
'canvas_width': 3,
'canvas_height': 1,
'category': 0,
'title': 'Beauty Blog Ad Woman with Unusual Hairstyle',
'type': [1, 3, 3, 3, 3, 4, 4, 4],
'left': [0.0,
-0.0009259259095415473,
0.24444444477558136,
0.5712962746620178,
0.2657407522201538,
0.369228333234787,
0.2739444375038147,
0.44776931405067444],
'top': [0.0,
-0.0009259259095415473,
0.37037035822868347,
0.41296297311782837,
0.41296297311782837,
0.8946287035942078,
0.4549448788166046,
0.40591198205947876],
'width': [1.0,
1.0018517971038818,
0.510185182094574,
0.16296295821666718,
0.16296295821666718,
0.30000001192092896,
0.4990740716457367,
0.11388888955116272],
'height': [1.0,
1.0018517971038818,
0.25833332538604736,
0.004629629664123058,
0.004629629664123058,
0.016611294820904732,
0.12458471953868866,
0.02657807245850563],
'opacity': [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0],
'text': ['', '', '', '', '', 'STAY WITH US', 'FOLLOW', 'PRESS'],
'font': [0, 0, 0, 0, 0, 152, 172, 152],
'font_size': [0.0, 0.0, 0.0, 0.0, 0.0, 18.0, 135.0, 30.0],
'text_align': [0, 0, 0, 0, 0, 2, 2, 2],
'angle': [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0],
'capitalize': [0, 0, 0, 0, 0, 0, 0, 0],
'line_height': [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0],
'letter_spacing': [0.0, 0.0, 0.0, 0.0, 0.0, 14.0, 12.55813980102539, 3.0],
'suitability': [0],
'keywords': ['beautiful',
'beauty',
'blog',
'blogging',
'caucasian',
'cute',
'elegance',
'elegant',
'fashion',
'fashionable',
'femininity',
'glamour',
'hairstyle',
'luxury',
'model',
'stylish',
'vogue',
'website',
'woman',
'post',
'instagram',
'ig',
'insta',
'fashion',
'purple'],
'industries': [1, 8, 13],
'color': [[153.0, 118.0, 96.0],
[34.0, 23.0, 61.0],
[34.0, 23.0, 61.0],
[255.0, 255.0, 255.0],
[255.0, 255.0, 255.0],
[255.0, 255.0, 255.0],
[255.0, 255.0, 255.0],
[255.0, 255.0, 255.0]],
'image': [<PIL.PngImagePlugin.PngImageFile image mode=RGBA size=256x256>,
<PIL.PngImagePlugin.PngImageFile image mode=RGBA size=256x256>,
<PIL.PngImagePlugin.PngImageFile image mode=RGBA size=256x256>,
<PIL.PngImagePlugin.PngImageFile image mode=RGBA size=256x256>,
<PIL.PngImagePlugin.PngImageFile image mode=RGBA size=256x256>,
<PIL.PngImagePlugin.PngImageFile image mode=RGBA size=256x256>,
<PIL.PngImagePlugin.PngImageFile image mode=RGBA size=256x256>,
<PIL.PngImagePlugin.PngImageFile image mode=RGBA size=256x256>]}
To get a label for categorical values, use the int2str method:
key = "font"
example = dataset[0]
dataset.features[key].int2str(example[key])
Data Fields
In the following, categorical fields are shown as categorical type, but the actual storage is int64.
Canvas attributes
| Field | Type | Shape | Description |
|---|---|---|---|
| id | string | () | Template ID from crello.com |
| group | categorical | () | Broad design groups, such as social media posts or blog headers |
| format | categorical | () | Detailed design formats, such as Instagram post or postcard |
| category | categorical | () | Topic category of the design, such as holiday celebration |
| canvas_width | categorical | () | Canvas pixel width |
| canvas_height | categorical | () | Canvas pixel height |
| length | int64 | () | Length of elements |
| suitability | categorical | (None,) | List of display tags, only mobile tag exists |
| keywords | string | (None,) | List of keywords associated to this template |
| industries | categorical | (None,) | List of industry tags like marketingAds |
Element attributes
| Field | Type | Shape | Description |
|---|---|---|---|
| type | categorical | (None,) | Element type, such as vector shape, image, or text |
| left | float32 | (None,) | Element left position normalized to [0, 1] range w.r.t. canvas_width |
| top | float32 | (None,) | Element top position normalized to [0, 1] range w.r.t. canvas_height |
| width | float32 | (None,) | Element width normalized to [0, 1] range w.r.t. canvas_width |
| height | float32 | (None,) | Element height normalized to [0, 1] range w.r.t. canvas_height |
| color | int64 | (None, 3) | Extracted main RGB color of the element |
| opacity | float32 | (None,) | Opacity in [0, 1] range |
| image | image | (None,) | Pre-rendered 256x256 preview of the element encoded in PNG format |
| text | string | (None,) | Text content in UTF-8 encoding for text element |
| font | categorical | (None,) | Font family name for text element |
| font_size | float32 | (None,) | Font size (height) in pixels |
| text_align | categorical | (None,) | Horizontal text alignment, left, center, right for text element |
| angle | float32 | (None,) | Element rotation angle (radian) w.r.t. the center of the element |
| capitalize | categorical | (None,) | Binary flag to capitalize letters |
| line_height | float32 | (None,) | Scaling parameter to line height, default is 1.0 |
| letter_spacing | float32 | (None,) | Adjustment parameter for letter spacing, default is 0.0 |
Note that the color and pre-rendered images do not necessarily accurately reproduce the original design templates. The original template is accessible at the following URL if still available.
https://create.vista.com/artboard/?template=<template_id>
left and top can be negative because elements can be bigger than the canvas size.
Data Splits
The Crello dataset has 3 splits: train, validation, and test. The current split is generated such that the same title of the original template shows up in only in one split.
| Split | Count |
|---|---|
| train | 18659 |
| validaton | 2391 |
| test | 2371 |
Visualization
Each example can be visualized in the following approach using skia-python. Note the following does not guarantee a similar appearance to the original template. Currently, the quality of text rendering is far from perfect.
import io
from typing import Any, Dict
import numpy as np
import skia
def render(features: datasets.Features, example: Dict[str, Any], max_size: float=512.) -> bytes:
"""Render parsed sequence example onto an image and return as PNG bytes."""
canvas_width = int(features["canvas_width"].int2str(example["canvas_width"]))
canvas_height = int(features["canvas_height"].int2str(example["canvas_height"]))
scale = min(1.0, max_size / canvas_width, max_size / canvas_height)
surface = skia.Surface(int(scale * canvas_width), int(scale * canvas_height))
with surface as canvas:
canvas.scale(scale, scale)
for index in range(example["length"]):
pil_image = example["image"][index]
image = skia.Image.frombytes(
pil_image.convert('RGBA').tobytes(),
pil_image.size,
skia.kRGBA_8888_ColorType)
left = example["left"][index] * canvas_width
top = example["top"][index] * canvas_height
width = example["width"][index] * canvas_width
height = example["height"][index] * canvas_height
rect = skia.Rect.MakeXYWH(left, top, width, height)
paint = skia.Paint(Alphaf=example["opacity"][index], AntiAlias=True)
angle = example["angle"][index]
with skia.AutoCanvasRestore(canvas):
if angle != 0:
degree = 180. * angle / np.pi
canvas.rotate(degree, left + width / 2., top + height / 2.)
canvas.drawImageRect(image, rect, paint=paint)
image = surface.makeImageSnapshot()
with io.BytesIO() as f:
image.save(f, skia.kPNG)
return f.getvalue()
Dataset Creation
Curation Rationale
The Crello dataset is compiled for the general study of vector graphic documents, with the goal of producing a dataset that offers complete vector graphic information suitable for neural methodologies.
Source Data
Initial Data Collection and Normalization
The dataset is initially scraped from the former crello.com and pre-processed to the above format.
Who are the source language producers?
While create.vista.com owns those templates, the templates seem to be originally created by a specific group of design studios.
Personal and Sensitive Information
The dataset does not contain any personal information about the creator but may contain a picture of people in the design template.
Considerations for Using the Data
Social Impact of Dataset
This dataset was developed for advancing the general study of vector graphic documents, especially for generative systems of graphic design. Successful utilization might enable the automation of creative workflow that human designers get involved in.
Discussion of Biases
The templates contained in the dataset reflect the biases appearing in the source data, which could present gender biases in specific design categories.
Other Known Limitations
Due to the unknown data specification of the source data, the color and pre-rendered images do not necessarily accurately reproduce the original design templates. The original template is accessible at the following URL if still available.
https://create.vista.com/artboard/?template=<template_id>
Additional Information
Dataset Curators
The Crello dataset was developed by Kota Yamaguchi.
Licensing Information
The origin of the dataset is create.vista.com (formally, crello.com).
The distributor ("We") do not own the copyrights of the original design templates.
By using the Crello dataset, the user of this dataset ("You") must agree to the
VistaCreate License Agreements.
The dataset is distributed under CDLA-Permissive-2.0 license.
Note
We do not re-distribute the original files as we are not allowed by terms.
Citation Information
@article{yamaguchi2021canvasvae,
title={CanvasVAE: Learning to Generate Vector Graphic Documents},
author={Yamaguchi, Kota},
journal={ICCV},
year={2021}
}
Releases
3.1: v3 release (Feb 16, 2023)
- Fix a bug that ignores newline characters in some of the texts
3.0: v3 release (Feb 13, 2023)
- Migrate to Hugging Face Hub.
- Fix various text rendering bugs.
- Change split generation criteria for avoiding near-duplicates: no compatibility with v2 splits.
- Incorporate a motion picture thumbnail in templates.
- Add
title,keywords,suitability, andindustriescanvas attributes. - Add
capitalize,line_height, andletter_spacingelement attributes.
2.0: v2 release (May 26, 2022)
- Add
text,font,font_size,text_align, andangleelement attributes. - Include rendered text element in
image_bytes.
1.0: v1 release (Aug 24, 2021)
Contributions
Thanks to @kyamagu for adding this dataset.