Pictograph Research
Dataset

FoodSeg103

Pictograph Research · Updated 2mo ago · Custom license

Dataset Card

FoodSeg103

A public computer-vision dataset of 7,118 images with 40,275 annotations across 103 classes, annotated as polygon.

Pictograph Research · a custom license · Public · Updated 2026-08-08

Overview

This dataset holds 7,118 images annotated as polygon across 103 classes. The most common auto-detected tags are eating, cooking, plate, bread, salad, sandwich. It is ready to fork, export in a dozen formats, or train a model from.

At a glance

Metric Value
Images7,118
Annotations40,275
Classes103
Annotation typespolygon
Folders2
StorageStandard
VisibilityPublic

What's in these images

Tags auto-detected across the dataset, most common first.

Tag Category Images
eatingAttributes5,652
cookingAttributes3,110
plateObjects2,615
breadObjects881
saladObjects833
sandwichObjects740
cakeObjects668
broccoliObjects533
fishObjects533
strawberryObjects398
burgerObjects394
carrotObjects380
pastaObjects350
cornObjects292
top-down viewAttributes281

Use this dataset

from pictograph import Client, Dataset

client = Client()  # reads PICTOGRAPH_API_KEY
dataset: Dataset = client.datasets.get("FoodSeg103")
print(dataset.image_count, dataset.annotation_count)
  • Fork it into your own workspace with one click from its page.
  • Pull it with the Pictograph SDK (shown above) or the CLI.
  • Export to COCO, YOLO, Darwin V7, Pascal VOC, and more.
  • Train a model directly from an export.

License

Released under a custom license.

7,118 images
Loading images…
Custom License Other
Annotations and benchmark code: Apache License 2.0 (the FoodSeg103 authors). Images: curated from the Recipe1M dataset (MIT CSAIL, im2recipe) and subject to Recipe1M terms of use - non-commercial research/educational use, registration-gated, redistribution not granted. Attribute both FoodSeg103 (Wu et al., ACM MM 2021) and Recipe1M (Marin et al.).

Join the discussion

Comments and reactions live in the app.

Sign in to comment

Start building today

$5/mo free compute. No payment method required.