---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: image_id
    dtype: int64
  - name: file_name
    dtype: string
  - name: width
    dtype: int32
  - name: height
    dtype: int32
  - name: objects
    struct:
    - name: bbox
      list:
        list: float32
        length: 4
    - name: category
      list:
        class_label:
          names:
            '0': ant
            '1': camel
            '2': jewellery
            '3': frog
            '4': physalis
            '5': celery
            '6': cauliflower
            '7': pepper
            '8': ranunculus
            '9': chess flower
            '10': cigarette
            '11': matthiola
            '12': cabbage
            '13': earring
            '14': dandelion
            '15': neroli
            '16': dragonfly
            '17': hyacinth
            '18': reptile/amphibia
            '19': apricot
            '20': snake
            '21': lizard
            '22': asparagus
            '23': spring onion
            '24': snowflake
            '25': moth
            '26': poppy
            '27': columbine
            '28': rabbit
            '29': geranium
            '30': crab
            '31': radish
            '32': big cat
            '33': jan steen jug
            '34': monkey
            '35': snail
            '36': bellflower
            '37': lilac
            '38': pot
            '39': peony
            '40': coffeepot
            '41': hazelnut
            '42': censer
            '43': artichoke
            '44': dahlia
            '45': sniffing
            '46': fly
            '47': deer
            '48': caterpillar
            '49': garlic
            '50': blackberry
            '51': chalice
            '52': lobster
            '53': necklace
            '54': bug
            '55': insect
            '56': prawn
            '57': bracelet
            '58': carrot
            '59': cornflower
            '60': pumpkin
            '61': orange
            '62': walnut
            '63': cat
            '64': daisy
            '65': forget-me-not
            '66': carafe
            '67': match
            '68': beer stein
            '69': tobacco-box
            '70': violet
            '71': pomander
            '72': bottle
            '73': candle
            '74': heliotrope
            '75': wine bottle
            '76': strawberry
            '77': pomegranate
            '78': whale
            '79': lily of the valley
            '80': iris
            '81': tobacco
            '82': olive
            '83': tobacco-packaging
            '84': meat
            '85': daffodil
            '86': melon
            '87': fire
            '88': petunia
            '89': mushroom
            '90': teapot
            '91': ring
            '92': pig
            '93': ashtray
            '94': cheese
            '95': onion
            '96': cup
            '97': nut
            '98': fig
            '99': drinking vessel
            '100': donkey
            '101': holding the nose
            '102': lily
            '103': smoke
            '104': bread
            '105': currant
            '106': glass without stem
            '107': anemone
            '108': mammal
            '109': chimney
            '110': smoking equipment
            '111': bivalve
            '112': butterfly
            '113': gloves
            '114': lemon
            '115': horse
            '116': plum
            '117': jasmine
            '118': pear
            '119': glass with stem
            '120': vegetable
            '121': carnation
            '122': jug
            '123': goat
            '124': fish
            '125': apple
            '126': tulip
            '127': cherry
            '128': cow
            '129': animal corpse
            '130': dog
            '131': fruit
            '132': bird
            '133': rose
            '134': peach
            '135': sheep
            '136': pipe
            '137': grapes
            '138': flower
    - name: area
      list: float32
    - name: iscrowd
      list: int64
  - name: artist
    dtype: string
  - name: title
    dtype: string
  - name: iconography
    dtype: string
  - name: earliest_date
    dtype: string
  - name: latest_date
    dtype: string
  - name: genre
    dtype: string
  - name: material
    dtype: string
  - name: photo_archive
    dtype: string
  - name: image_credits
    dtype: string
  - name: details_url
    dtype: string
  - name: iconclass_code
    dtype: string
  - name: image_license
    dtype: string
  - name: description
    dtype: string
  - name: keywords
    dtype: string
  - name: language
    dtype: string
  splits:
  - name: train
    num_bytes: 1335652730
    num_examples: 4264
  - name: test
    num_bytes: 74738280
    num_examples: 448
  download_size: 738282481
  dataset_size: 1410391010
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
license: cc-by-4.0
task_categories:
- object-detection
tags:
- art
- artwork
- object-detection
- digital-humanities
- iconclass
- glam
- cultural-heritage
size_categories:
- 1K<n<10K
---

# ODOR — Object Detection for Olfactory References in Artworks

4,712 artwork images with **38,165 bounding-box annotations** across **139 fine-grained
categories** of smell-related objects — flowers, fruit, censers, animals, vessels — drawn from
European art.

Computer vision on artworks is hard in ways photographic benchmarks are not: artistic
abstraction, peripheral objects, and fine-grained distinctions between visually similar
classes. ODOR is built to test exactly that.

## What's in a row

Each row carries the image, its detections, **and its art-historical metadata** — the Zenodo
deposit ships these as separate files, and they are joined here:

- `image`, `image_id`, `file_name`, `width`, `height`
- `objects` — `bbox` (COCO `xywh`), `category` (`ClassLabel`, 139 names), `area`, `iscrowd`
- `artist`, `title`, `iconography`, **`iconclass_code`**, `genre`, `material`
- `earliest_date`, `latest_date`, `photo_archive`, `image_credits`, `details_url`,
  `description`, `keywords`, `language`

Where present, `iconclass_code` connects this dataset to
[`biglam/brill_iconclass`](https://huggingface.co/datasets/biglam/brill_iconclass) — but note it is
sparse (see caveats).

## Splits

| Split | Images |
|---|---|
| train | 4,264 |
| test | 448 |

## Caveats

- **Metadata coverage is uneven.** `artist` is present on 96.6% of rows, but **`iconclass_code` on only 541 of 4,264 (12.7%)**. Do not assume Iconclass coverage.
- Metadata is **multilingual** (`language` is `de` for much of the Städel material) and not normalised across source archives.
- `earliest_date` / `latest_date` are inconsistent in the source — some are years, some centuries (e.g. `"18"`). Left as strings rather than guessed at.
- Images come from several photo archives with differing terms; `image_credits` and `details_url` point back to the originals.

## Source & credit

Mathias Zinnen, Prathmesh Madhu, Ronak Kosti et al. *The Object Detection for Olfactory
References (ODOR) Dataset* (v3.0.2). Zenodo, 2024-04-26.
<https://zenodo.org/records/11070878> — CC-BY-4.0.

Produced in the context of the **Odeuropa** project. This repository joins the deposit's COCO
annotations with its `meta.csv` and converts to Parquet. Please cite the original authors.