lang
stringclasses
2 values
audio
audioduration (s)
0.1
51.3
image
imagewidth (px)
30
4.9k
text
stringlengths
7
60
cls
stringclasses
1k values
cls_id
int64
0
999
target_text
stringlengths
2
35
el
swab_πλατύγυρος πίλος
sombrero
808
πλατύγυρος πίλος
en
Ιαπωνικό σπάνιελ_jaguar
jaguar
290
jaguar
en
σκληρός δίσκος_sock
sock
806
sock
el
dumbbell_φυσητήρας χειρός
hand blower
589
φυσητήρας χειρός
el
τεριέ παιχνιδιών_agama
toy terrier
158
τεριέ παιχνιδιών
el
δενδρόβιο ζώο της αυστραλίας_convertible
koala
105
δενδρόβιο ζώο της αυστραλίας
en
μικρόφωνο_stove
stove
827
stove
el
crossword puzzle_βλήμα
missile
657
βλήμα
el
little blue heron_μέλισσα
bee
309
μέλισσα
el
house finch_θαλάσσιο λιοντάρι
sea lion
150
θαλάσσιο λιοντάρι
en
κανάτα νερού_red-backed sandpiper
red-backed sandpiper
140
red-backed sandpiper
el
πιατοθήκη_flute
plate rack
729
πιατοθήκη
en
diaper_αγκαθωτός αστακός
diaper
529
diaper
en
turnstile_σαχλαμάρα
turnstile
877
turnstile
en
κράνος συντριβής_volleyball
volleyball
890
volleyball
el
organ_όμορφη
hay
958
όμορφη
el
komondor_μαράκα
maraca
641
μαράκα
en
great grey owl_πουλί γερανού
great grey owl
24
great grey owl
en
abaya_σαύρα φουρνωτή
abaya
399
abaya
el
china cabinet_βιβλίο κόμικ
comic book
917
βιβλίο κόμικ
el
Ντόμπερμαν_rifle
Doberman
236
Ντόμπερμαν
el
pug_λαγωνικό
bloodhound
163
λαγωνικό
en
trifle_αχυροσκεπή
trifle
927
trifle
en
flatworm_μαλαμούτε της αλάσκας
flatworm
110
flatworm
el
δοχείο_jersey
pot
738
δοχείο
el
vending machine_λουκάνικο
hotdog
934
λουκάνικο
el
μπέιζμπολ_Shetland sheepdog
baseball
429
μπέιζμπολ
el
Mexican hairless_νερόφιδο
water snake
58
νερόφιδο
en
σκύλος Samoyed_Brittany spaniel
Brittany spaniel
215
Brittany spaniel
en
μικρό λεωφορείο_vending machine
vending machine
886
vending machine
el
malamute_Νορβηγικό ελκχούν
Norwegian elkhound
174
Νορβηγικό ελκχούν
el
cabbage butterfly_συσκευή CD
CD player
485
συσκευή CD
el
timber wolf_Αφρικανικός ελέφαντας
African elephant
386
Αφρικανικός ελέφαντας
el
washer_βόλεϊ
volleyball
890
βόλεϊ
en
snorkel_ρόδι
snorkel
801
snorkel
el
marimba_Θιβετιανό Μαστίφ
Tibetan mastiff
244
Θιβετιανό Μαστίφ
en
κασέτα_guillotine
guillotine
583
guillotine
en
ντουλάπι Κίνας_bath towel
bath towel
434
bath towel
el
ballplayer_ρακέτα
racket
752
ρακέτα
el
hammerhead_γκρίζα αλεπού
grey fox
280
γκρίζα αλεπού
el
butternut squash_Γραβάτα Windsor
Windsor tie
906
Γραβάτα Windsor
el
giant panda_σκορπιός
scorpion
71
σκορπιός
el
ραδιόφωνο_Labrador retriever
radio
754
ραδιόφωνο
el
ηλεκτρική ατμομηχανή_snowmobile
electric locomotive
547
ηλεκτρική ατμομηχανή
el
ηλιακό πιάτο_snail
solar dish
807
ηλιακό πιάτο
el
whiskey jug_Σαλούκι
Saluki
176
Σαλούκι
el
plastic bag_κίτρινη γυναικεία παντόφλα
yellow lady's slipper
986
κίτρινη γυναικεία παντόφλα
en
μπαλοπαίκτης_trench coat
trench coat
869
trench coat
el
ποτηρι ζεσεως_nematode
beaker
438
ποτηρι ζεσεως
en
πράσινο φίδι_house finch
house finch
12
house finch
en
turnstile_αυλαία του θεάτρου
turnstile
877
turnstile
en
βάθρο_hare
hare
331
hare
el
καϊκι_racer
boathouse
449
καϊκι
el
στήθος φαρμάκου_folding chair
medicine chest
648
στήθος φαρμάκου
en
κρήνη_ruffed grouse
ruffed grouse
82
ruffed grouse
el
upright_κράνος pickelhaube
pickelhaube
715
κράνος pickelhaube
el
komondor_Λευκό τεριέ West Highland
West Highland white terrier
203
Λευκό τεριέ West Highland
el
king snake_άμαξα
horse cart
603
άμαξα
el
hay_μαστιγγιοουρά
whiptail
41
μαστιγγιοουρά
el
trailer truck_πύργος νερού
water tower
900
πύργος νερού
en
knee pad_δακτυλήθρα
knee pad
615
knee pad
el
κερί_shoe shop
candle
470
κερί
el
gas pump_Αγγλικό ελατήριο
English springer
217
Αγγλικό ελατήριο
el
ορεινός λέων_guillotine
cougar
286
ορεινός λέων
en
ελαιοκράμβη_maraca
maraca
641
maraca
el
doormat_στήριγμα λαιμού
neck brace
678
στήριγμα λαιμού
el
miniature pinscher_Μαλινουά
malinois
225
Μαλινουά
en
θαλάσσιο θηλαστικό dugong_echidna
echidna
102
echidna
en
chiton_Σκύλος Greater Swiss Mountain
chiton
116
chiton
en
τηλεόραση_red wolf
red wolf
271
red wolf
en
είδος αιγίθαλου_jaguar
jaguar
290
jaguar
en
Indian cobra_κεφάλι λάχανο
Indian cobra
63
Indian cobra
en
Sealyham terrier_διαμάντι
Sealyham terrier
190
Sealyham terrier
el
σκαθάρι τίγρης_miniskirt
tiger beetle
300
σκαθάρι τίγρης
el
παζλ_snail
jigsaw puzzle
611
παζλ
en
είδος άγριας όρνιθος_pineapple
pineapple
953
pineapple
el
πράσινο mamba_hand blower
green mamba
64
πράσινο mamba
el
lotion_στάμνα
pitcher
725
στάμνα
en
box turtle_κινούμενος από
box turtle
37
box turtle
en
μπουκάλι κρασιού_military uniform
military uniform
652
military uniform
el
καρμανιόλα_capuchin
guillotine
583
καρμανιόλα
en
carbonara_μπαλτάς
carbonara
959
carbonara
en
κόκκινο κόκκαλο_American Staffordshire terrier
American Staffordshire terrier
180
American Staffordshire terrier
en
hammer_κέντρο προσοχής
hammer
587
hammer
en
σκύλος Samoyed_eel
eel
390
eel
el
toy poodle_Ιαπωνικό σπάνιελ
Japanese spaniel
152
Ιαπωνικό σπάνιελ
en
είδος πιθήκου_radiator
radiator
753
radiator
el
analog clock_χέλι
eel
390
χέλι
el
κοινός πόλεκας_pomegranate
polecat
358
κοινός πόλεκας
el
snowplow_γορίλλας
gorilla
366
γορίλλας
el
damselfly_πετσέτα μπάνιου
bath towel
434
πετσέτα μπάνιου
en
ricksha_racer
racer
751
racer
el
little blue heron_σκύλος τεριέ συνόρων
Border terrier
182
σκύλος τεριέ συνόρων
en
φορείο_junco
junco
13
junco
el
κότα-του-δάσους_moving van
hen-of-the-woods
996
κότα-του-δάσους
en
κουνέλι ανγκόρα_armadillo
armadillo
363
armadillo
el
Μέσης_EntleBucher
mitten
658
Μέσης
el
διαστημικό λεωφορείο_white wolf
space shuttle
812
διαστημικό λεωφορείο
en
τσιμπούρι_goose
goose
99
goose
el
breakwater_μονόκυκλο
unicycle
880
μονόκυκλο

Dataset Card for Symile-M3

Symile-M3 is a multilingual dataset of (audio, image, text) samples. The dataset is specifically designed to test a model's ability to capture higher-order information between three distinct high-dimensional data types: by incorporating multiple languages, we construct a task where text and audio are both needed to predict the image, and where, importantly, neither text nor audio alone would suffice.

Overview

image/jpeg

Let w represent the number of languages in the dataset (w=2, w=5, and w=10 correspond to Symile-M3-2, Symile-M3-5, and Symile-M3-10, respectively). An (audio, image, text) sample is generated by first drawing a short one-sentence audio clip from Common Voice spoken in one of w languages with equal probability. An image is drawn from ImageNet that corresponds to one of 1,000 classes with equal probability. Finally, text containing exactly w words is generated based on the drawn audio and image: one of the w words in the text is the drawn image class name in the drawn audio language. The remaining w-1 words are randomly chosen from the ImageNet class names and written in one of the w languages such that there is no overlap in language or class name across the w words in the text. The words are separated by underscores, and their order is randomized.

Tasks

The dataset was designed to evaluate a model on the zero-shot retrieval task of finding an image of the appropriate class given the audio and text. The most probable image for a given query audio and text pair, selected from all possible candidate images in the test set, is that with the highest similarity score.

The dataset was designed to ensure that neither text nor audio alone would suffice to predict the image. Therefore, success on this zero-shot retrieval task hinges on a model's ability to capture joint information between the three modalities.

Dataset Structure

Each sample in the dataset is a dictionary containing the following fields:

{
    # language code of the audio clip
    'lang': 'ja',
    
    # audio data
    'audio': {
        'path': 'common_voice_ja_39019065.mp3',                 # Common Voice filename
        'array': array([0.00000000e+00, ..., 7.78421963e-06]),  # raw audio waveform
        'sampling_rate': 32000                                  # sampling rate in Hz
    },
    
    # image as a PIL Image object (RGB, size varies)
    'image': <PIL.JpegImageFile image mode=RGB size=500x375>,
    
    # text containing w words (one per language) separated by underscores
    'text': 'σπιτάκι πουλιών_ドーム_प्रयोगशाला कोट_мавпа-павук_gown',
    
    # target word class name in English (key in translations.json)
    'cls': 'dome',
    
    # class ID from translations.json (0 to 999)
    'cls_id': 538,
    
    # target word (class name in the language of the audio)
    'target_text': 'ドーム'
}

The dataset includes a translations.json file that maps ImageNet class names across all supported languages. Each entry contains:

  • The English class name as the key
  • Translations for all supported languages (ar, el, en, hi, ja, ko, te, th, uk, zh-CN)
  • The ImageNet synset ID
  • A unique class ID (0-999)

Example structure:

{
    "tench": {
        "synset_id": "n01440764",
        "cls_id": 0,
        "ar": "سمك البنش",
        "el": "είδος κυπρίνου",
        "en": "tench",
        "hi": "टेंच",
        "ja": "テンチ",
        "ko": "텐치",
        "te": "టెంచ్",
        "th": "ปลาเทนช์",
        "uk": "линь",
        "zh-CN": "丁鱥"
    }
}

Dataset Variants

We release three variants of the dataset:

  • Symile-M3-2 with 2 languages: English (en) and Greek (el).
  • Symile-M3-5 with 5 languages: English (en), Greek (el), Hindi (hi), Japanese (ja), and Ukrainian (uk).
  • Symile-M3-10 with 10 languages: Arabic (ar), Greek (el), English (en), Hindi (hi), Japanese (ja), Korean (ko), Telugu (te), Thai (th), Ukrainian (uk), and Chinese (zh-CN).

Each variant is available in four sizes:

  • Large (l): 10M training samples, 500K validation samples, 500K test samples
  • Medium (m): 5M training samples, 250K validation samples, 250K test samples
  • Small (s): 1M training samples, 50K validation samples, 50K test samples
  • Extra Small (xs): 500K training samples, 25K validation samples, 25K test samples

Usage

Before using the dataset, ensure you have the required audio and image processing libraries installed:

pip install librosa soundfile pillow

To load a specific version of Symile-M3, use a configuration name following the pattern symile-m3-{num_langs}-{size} where:

  • num_langs is 2, 5, or 10
  • size is xs, s, m, or l

For example, to load the xs version of Symile-M3-5:

from datasets import load_dataset

dataset = load_dataset("arsaporta/symile-m3", "symile-m3-5-xs")

print(dataset['train'][0])    # access first train sample
print(len(dataset['train']))  # get number of train samples

To process the dataset without loading it entirely into memory, use streaming mode to load samples one at a time:

from datasets import load_dataset

dataset = load_dataset("arsaporta/symile-m3", "symile-m3-5-xs", streaming=True)

print(next(iter(dataset['train'])))

To download the dataset for offline use:

import os
from datasets import load_dataset
from huggingface_hub import snapshot_download

local_dir = "./symile-m3-5-xs" # where to save

# download parquet files
snapshot_download(
    repo_id="arsaporta/symile-m3",
    repo_type="dataset",
    local_dir=local_dir,
    allow_patterns=["symile-m3-5-xs/*"]  # which configuration to download
)

# load the downloaded parquet files
dataset = load_dataset(
    "parquet",
    data_files={
        "train": os.path.join(data_dir, "train-*.parquet"),
        "validation": os.path.join(data_dir, "val-*.parquet"),
        "test": os.path.join(data_dir, "test-*.parquet")
    }
)

Working with Raw Data

To work directly with the source images (jpeg) and audio (mp3):

  1. Download the source data:

    • ImageNet: Get the training data from Kaggle's ImageNet Challenge
    • Common Voice: Download your needed languages from Common Voice:
      • All languages use Common Voice v16.0, except English which uses v14.0
      • Required languages vary by configuration:
        • Symile-M3-2: English (en), Greek (el)
        • Symile-M3-5: English, Greek, Hindi (hi), Japanese (ja), Ukrainian (uk)
        • Symile-M3-10: All of the above plus Arabic (ar), Korean (ko), Telugu (te), Thai (th), Chinese (zh-CN)
  2. Access the dataset CSV files:

    • Find them in the .csv_files directory, organized by configuration (e.g., symile-m3-2-xs, symile-m3-10-l)
    • Each configuration contains train.csv, val.csv, and test.csv
    • CSV paths match the default extraction paths of ImageNet (ILSVRC/Data/CLS-LOC/train/...) and Common Voice (cv/{lang}/clips/...)

Citation

@inproceedings{saporta2024symile,
  title = {Contrasting with Symile: Simple Model-Agnostic Representation Learning for Unlimited Modalities}
  author = {Saporta, Adriel and Puli, Aahlad and Goldstein, Mark and Ranganath, Rajesh}
  booktitle = {Advances in Neural Information Processing Systems},
  year = {2024}
}
Downloads last month
131
Edit dataset card