datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
iNat-2021-traininatspectro-bioacoustics
iNatSpectro Bioacoustics
A community-contributed dataset of bioacoustic call annotations pointing at iNaturalist audio observations. Contributions are made via the iNatSpectro browser extension: a user listens to and views the spectrogram of an iNaturalist recording, marks a call, and submits a pointer to the source recording plus the marked time/frequency bounding box. Each record carries the observation's quality_grade, so consumers can scope training to research-grade data… See the full description on the dataset page: https://huggingface.co/datasets/japht/inatspectro-bioacoustics.inaturalist-s3-massiveSnapshot of the iNaturalist dataset from 2026-03-27.
NOTE This dataset only contains images with license CC0 or CC-BY.
The CSV files were retrieved using
aws s3 --no-sign-request --region eu-central-1 cp s3://inaturalist-open-data/photos.csv.gz photos.csv.gz
aws s3 --no-sign-request --region eu-central-1 cp s3://inaturalist-open-data/taxa.csv.gz taxa.csv.gz
aws s3 --no-sign-request --region eu-central-1 cp s3://inaturalist-open-data/observations.csv.gz observations.csv.gz
then this dataset… See the full description on the dataset page: https://huggingface.co/datasets/philipp-zettl/inaturalist-s3-massive.inaturalist-enriched
Enriched iNaturalist dataset from 2026-03-27.
This dataset is based on philipp-zettl/inaturalist-s3-massive.
The data was enriched using the ./enrich.py script inside the repository.
It contains the following features
photo_id: The original ID of the photo inside the inaturalist dataset
observation_uuid: The observation's UUID
image: The actual image content
taxon_id: The ID of the taxonomy
species_name: The name of the species inside the image
taxonomic_rank: The type of taxonomic rank… See the full description on the dataset page: https://huggingface.co/datasets/philipp-zettl/inaturalist-enriched.wds_inaturalistiNat-2021-valinat_soundsiNatinat_soundsinat_butterfliesThis dataset contains images from iNaturalist of butterflies (superfamily Papilionoidea) with at least one fave. Check the descriptions - some images have a licence like CC-BY-NC and can't be used for commercial purposes.
The list of observations was exported from iNaturalist after a query similar to https://www.inaturalist.org/observations?place_id=any&popular&taxon_id=47224
The images were downloaded with img2dataset and uploaded to the huggingface hub by @johnowhitaker using this colab… See the full description on the dataset page: https://huggingface.co/datasets/huggan/inat_butterflies.inat-species-explorer-dataiNat21_filtered
iNat21 manually filtered masks
This repository contains the original iNat21 mask proposals, the manual review decisions, and the filtered mask parquet produced from those decisions.
The masks were reviewed with the ConCor Wildlife Verification Interface, a CPU-only browser interface for stepping through images and mask proposals, accepting or deleting each mask, saving review progress to JSON, and exporting a filtered parquet.
Files
File
Description
Rows /… See the full description on the dataset page: https://huggingface.co/datasets/suryadv/iNat21_filtered.inaturalistinaturalist_1kinat_soundsinaturalist
Dataset Card for inaturalist
This dataset is comprised of 9549 observations that were posted on the iNaturalist app. iNaturalist is a website and app that 'aims to provide a crowd sourced identification system' for plants, insects, and animals.
Dataset Details
For each of the 9549 observations included in this dataset, there is information about the quality ('quality_grade') of the observation and associated photo(s),
a species name ('species_guess') along with a… See the full description on the dataset page: https://huggingface.co/datasets/ba188/inaturalist.iNaturalist_v2
Dataset Card for Dataset Name
This dataset is comprised of 1,079 observations that were posted on the iNaturalist app. iNaturalist is a website and mobile app that 'aims to
provide a crowd-sourced identification system' for plants, insects, and animals.
Dataset Details
Dataset Description
For each of the 1,079 observations included in this dataset, there is information about the quality of the associated image (quality_grade), a species label… See the full description on the dataset page: https://huggingface.co/datasets/ba188/iNaturalist_v2.iNatIGCD
Dataset Card for Dataset Name
Dataset Summary
This dataset is built on the iNaturalist 2021 dataset and is used for the Incremental Generalized Category Discovery task.
For more information about the task, please checkout this paper.
Supported Tasks and Leaderboards
[More Information Needed]
Dataset Structure
Data Instances
[More Information Needed]
Data Fields
[More Information Needed]
Data Splits
[More Information… See the full description on the dataset page: https://huggingface.co/datasets/tennant/iNatIGCD.inaturalist-2024-2.8k-claude-opus-5-recaptioned
iNaturalist 2024 2.8K — Claude Opus 5 Recaptioned
This is a 2,824-image derivative subset of iNaturalist 2024 (iNat24), distributed through the INQUIRE project, selected through the inaturalist portion of zlab-princeton/i1-captions. It is not the complete 4.8-million-image iNat24 training set.
Every image has one newly generated, detailed English caption. The recaptioning was performed with Claude Opus 5 via Claude Code on August 2, 2026. The image was the primary evidence; the… See the full description on the dataset page: https://huggingface.co/datasets/sirus/inaturalist-2024-2.8k-claude-opus-5-recaptioned.bats_inat_priNat21-1shot-fewshotsThis is the official release of the training dataset for paper Taxonomy-Aware Representation Alignment for Hierarchical Visual Recognition with Large Multimodal Models. Code is available at https://github.com/PKU-ICST-MIPL/TARA_CVPR2026.
Data Source
Training
We randomly sample 1-shot data per category from iNaturalist2021 dataset.
Data Fields
image: input image(s)
data type: dict
problem: input question
data type: string
label: coarse-to-fine categories… See the full description on the dataset page: https://huggingface.co/datasets/StevenHH2000/iNat21-1shot-fewshots.inat_butterflies_top10kFiltered version of https://huggingface.co/datasets/huggan/inat_butterflies
To pick the best images, CLIP was used to compare each image with a text description of a good image ("")
Notebook for the filtering: https://colab.research.google.com/drive/1OEqr1TtL4YJhdj_bebNWXRuG3f2YqtQE?usp=sharing
See the original dataset for sources and licence caveats (tl;dr check the image descriptions to make sure you aren't breaking a licence like CC-BY-NC-ND which some images have)
INatSoundsMiniiNaturalist_mortality_records_12Apr2025inaturalist-taxonomy-multimodalinat24_tiny
The iNaturalist 2024 Dataset, Tiny (iNat24-tiny)
This dataset contains 1,000 images from iNat24, spanning 100 species classes with 10 images each.
inat2021-train-plants-meta
juppy44/inat2021-train-plants-meta
iNaturalist 2021 train split, plants-only metadata (no images).
Source: iNaturalist 2021 train images on local filesystem
Filter: kingdom = Plantae (parsed from class directory names like 08342_Plantae_Tracheophyta_...)
Total plant samples uploaded: 1000000
Stored as multiple Parquet shards under data/.
Columns:
filepath: full path of the image on the creator's machine
class_dir: the class directory name (encodes species/taxonomy)
kingdom:… See the full description on the dataset page: https://huggingface.co/datasets/juppy44/inat2021-train-plants-meta.inaturalist_valinaturalist_sorted_with_labelinat-2017-subset
iNaturalist 2017 Supercategory Subset
This dataset is a sampled subset of the iNaturalist 2017 Challenge dataset, specifically processed for efficient object detection training across major biological supercategories.
Dataset Summary
Source: iNaturalist 2017 (Competition Version)
Task: Object Detection
Classes: 9 (Collapsed from thousands of species into biological supercategories)
Training Samples: 1,000 images per supercategory (~9,000 total)
Validation Samples: 200… See the full description on the dataset page: https://huggingface.co/datasets/MacroSony/inat-2017-subset.