Datasets:
metadata
annotations_creators:
- derived
language:
- eng
license: unknown
multilinguality: monolingual
source_datasets:
- myang333/BioVITAA2IRetrieval
task_categories:
- other
task_ids: []
dataset_info:
- config_name: unseen_genus-corpus
features:
- name: id
dtype: string
- name: image
dtype: image
- name: taxon
dtype: string
splits:
- name: test
num_bytes: 3370275324
num_examples: 2835
download_size: 3393202400
dataset_size: 3370275324
- config_name: unseen_genus-qrels
features:
- name: query-id
dtype: string
- name: corpus-id
dtype: string
- name: score
dtype: int64
splits:
- name: test
num_bytes: 507202
num_examples: 11129
download_size: 29877
dataset_size: 507202
- config_name: unseen_genus-queries
features:
- name: audio
dtype: audio
- name: id
dtype: string
- name: correct_taxon
dtype: string
- name: candidate_taxa
list: string
splits:
- name: test
num_bytes: 1537422398
num_examples: 730
download_size: 1537466646
dataset_size: 1537422398
- config_name: unseen_genus-top_ranked
features:
- name: query-id
dtype: string
- name: corpus-ids
list: string
splits:
- name: test
num_bytes: 7115306
num_examples: 730
download_size: 7102616
dataset_size: 7115306
- config_name: unseen_species-corpus
features:
- name: id
dtype: string
- name: image
dtype: image
- name: taxon
dtype: string
splits:
- name: test
num_bytes: 3370303485
num_examples: 2835
download_size: 3393207374
dataset_size: 3370303485
- config_name: unseen_species-qrels
features:
- name: query-id
dtype: string
- name: corpus-id
dtype: string
- name: score
dtype: int64
splits:
- name: test
num_bytes: 501191
num_examples: 8724
download_size: 33887
dataset_size: 501191
- config_name: unseen_species-queries
features:
- name: audio
dtype: audio
- name: id
dtype: string
- name: correct_taxon
dtype: string
- name: candidate_taxa
list: string
splits:
- name: test
num_bytes: 1759479811
num_examples: 887
download_size: 1759520176
dataset_size: 1759479811
- config_name: unseen_species-top_ranked
features:
- name: query-id
dtype: string
- name: corpus-ids
list: string
splits:
- name: test
num_bytes: 6705319
num_examples: 887
download_size: 6681287
dataset_size: 6705319
configs:
- config_name: unseen_genus-corpus
data_files:
- split: test
path: unseen_genus-corpus/test-*
- config_name: unseen_genus-qrels
data_files:
- split: test
path: unseen_genus-qrels/test-*
- config_name: unseen_genus-queries
data_files:
- split: test
path: unseen_genus-queries/test-*
- config_name: unseen_genus-top_ranked
data_files:
- split: test
path: unseen_genus-top_ranked/test-*
- config_name: unseen_species-corpus
data_files:
- split: test
path: unseen_species-corpus/test-*
- config_name: unseen_species-qrels
data_files:
- split: test
path: unseen_species-qrels/test-*
- config_name: unseen_species-queries
data_files:
- split: test
path: unseen_species-queries/test-*
- config_name: unseen_species-top_ranked
data_files:
- split: test
path: unseen_species-top_ranked/test-*
tags:
- mteb
- audio
- image
Measures whether a model can connect an animal's call to its appearance without text as an intermediary. Each query is a field recording of a single animal, and the model ranks 100 candidate taxa -- the recorded taxon plus 99 distractors -- over an index of 2,835 wildlife photographs, where a taxon is represented by every photograph of that taxon. A taxon scores its best-matching photograph and the 100 taxa are ranked by that score, so the reported taxon_top_k_accuracy is taxon-level rather than document-level. Runs 887 queries at species level and 730 at genus level over BioVITA's held-out unseen split: 325 species excluded from training and the 225 genera they belong to.
| Task category | Any2AnyRetrieval (audio-to-image) |
| Domains | Bioacoustics, Nature, Encyclopaedic |
| Reference | CVPR |
Source datasets:
How to evaluate on this task
You can evaluate an embedding model on this dataset using the following code:
import mteb
task = mteb.get_task("BioVITAA2IRetrieval")
model = mteb.get_model(YOUR_MODEL)
mteb.evaluate(model, task)
To learn more about how to run models on mteb task check out the GitHub repository.
Citation
If you use this dataset, please cite the dataset as well as mteb, as this dataset likely includes additional processing as a part of the MMTEB Contribution.
@inproceedings{shinoda2026biovita,
author = {Shinoda, Risa and Shiohara, Kaede and Inoue, Nakamasa and Saito, Kuniaki and Santo, Hiroaki and Okura, Fumio},
booktitle = {CVPR},
title = {BioVITA: Biological Dataset, Model, and Benchmark for Visual-Textual-Acoustic Alignment},
year = {2026},
}
@article{enevoldsen2025mmtebmassivemultilingualtext,
title={MMTEB: Massive Multilingual Text Embedding Benchmark},
author={Kenneth Enevoldsen and Isaac Chung and Imene Kerboua and Márton Kardos and Ashwin Mathur and David Stap and Jay Gala and Wissam Siblini and Dominik Krzemiński and Genta Indra Winata and Saba Sturua and Saiteja Utpala and Mathieu Ciancone and Marion Schaeffer and Gabriel Sequeira and Diganta Misra and Shreeya Dhakal and Jonathan Rystrøm and Roman Solomatin and Ömer Çağatan and Akash Kundu and Martin Bernstorff and Shitao Xiao and Akshita Sukhlecha and Bhavish Pahwa and Rafał Poświata and Kranthi Kiran GV and Shawon Ashraf and Daniel Auras and Björn Plüster and Jan Philipp Harries and Loïc Magne and Isabelle Mohr and Mariya Hendriksen and Dawei Zhu and Hippolyte Gisserot-Boukhlef and Tom Aarsen and Jan Kostkan and Konrad Wojtasik and Taemin Lee and Marek Šuppa and Crystina Zhang and Roberta Rocca and Mohammed Hamdy and Andrianos Michail and John Yang and Manuel Faysse and Aleksei Vatolin and Nandan Thakur and Manan Dey and Dipam Vasani and Pranjal Chitale and Simone Tedeschi and Nguyen Tai and Artem Snegirev and Michael Günther and Mengzhou Xia and Weijia Shi and Xing Han Lù and Jordan Clive and Gayatri Krishnakumar and Anna Maksimova and Silvan Wehrli and Maria Tikhonova and Henil Panchal and Aleksandr Abramov and Malte Ostendorff and Zheng Liu and Simon Clematide and Lester James Miranda and Alena Fenogenova and Guangyu Song and Ruqiya Bin Safi and Wen-Ding Li and Alessia Borghini and Federico Cassano and Hongjin Su and Jimmy Lin and Howard Yen and Lasse Hansen and Sara Hooker and Chenghao Xiao and Vaibhav Adlakha and Orion Weller and Siva Reddy and Niklas Muennighoff},
publisher = {arXiv},
journal={arXiv preprint arXiv:2502.13595},
year={2025},
url={https://arxiv.org/abs/2502.13595},
doi = {10.48550/arXiv.2502.13595},
}
@article{muennighoff2022mteb,
author = {Muennighoff, Niklas and Tazi, Nouamane and Magne, Loïc and Reimers, Nils},
title = {MTEB: Massive Text Embedding Benchmark},
publisher = {arXiv},
journal={arXiv preprint arXiv:2210.07316},
year = {2022}
url = {https://arxiv.org/abs/2210.07316},
doi = {10.48550/ARXIV.2210.07316},
}
Dataset Statistics
Dataset Statistics
The following code contains the descriptive statistics from the task. These can also be obtained using:
import mteb
task = mteb.get_task("BioVITAA2IRetrieval")
desc_stats = task.metadata.descriptive_stats
{
"test": {
"num_samples": 7287,
"num_queries": 1617,
"num_documents": 5670,
"number_of_characters": 0,
"documents_text_statistics": null,
"documents_image_statistics": {
"min_image_width": 240,
"average_image_width": 1689.7855379188713,
"max_image_width": 2048,
"min_image_height": 143,
"average_image_height": 1347.004585537919,
"max_image_height": 2048,
"unique_images": 2813
},
"documents_audio_statistics": null,
"documents_video_statistics": null,
"queries_text_statistics": null,
"queries_image_statistics": null,
"queries_audio_statistics": {
"total_duration_seconds": 53711.96968331066,
"min_duration_seconds": 0.8950113378684807,
"average_duration_seconds": 33.217049896914446,
"max_duration_seconds": 615.0101979166667,
"unique_audios": 912,
"average_sampling_rate": 45184.44650587508,
"sampling_rates": {
"44100": 1202,
"48000": 370,
"24000": 4,
"22050": 21,
"200000": 1,
"32000": 3,
"16000": 6,
"96000": 6,
"192000": 4
}
},
"queries_video_statistics": null,
"relevant_docs_statistics": {
"num_relevant_docs": 19853,
"min_relevant_docs_per_query": 1,
"average_relevant_docs_per_query": 12.277674706246135,
"max_relevant_docs_per_query": 95,
"unique_relevant_docs": 5670
},
"top_ranked_statistics": {
"num_top_ranked": 1821030,
"min_top_ranked_per_query": 924,
"average_top_ranked_per_query": 1126.178107606679,
"max_top_ranked_per_query": 1509
},
"hf_subset_descriptive_stats": {
"unseen_species": {
"num_samples": 3722,
"num_queries": 887,
"num_documents": 2835,
"number_of_characters": 0,
"documents_text_statistics": null,
"documents_image_statistics": {
"min_image_width": 240,
"average_image_width": 1689.7855379188713,
"max_image_width": 2048,
"min_image_height": 143,
"average_image_height": 1347.004585537919,
"max_image_height": 2048,
"unique_images": 2813
},
"documents_audio_statistics": null,
"documents_video_statistics": null,
"queries_text_statistics": null,
"queries_image_statistics": null,
"queries_audio_statistics": {
"total_duration_seconds": 29194.276403194446,
"min_duration_seconds": 0.8950113378684807,
"average_duration_seconds": 32.91350214565327,
"max_duration_seconds": 615.0101979166667,
"unique_audios": 887,
"average_sampling_rate": 45242.16459977452,
"sampling_rates": {
"44100": 652,
"48000": 210,
"24000": 2,
"22050": 12,
"200000": 1,
"32000": 2,
"16000": 3,
"96000": 3,
"192000": 2
}
},
"queries_video_statistics": null,
"relevant_docs_statistics": {
"num_relevant_docs": 8724,
"min_relevant_docs_per_query": 1,
"average_relevant_docs_per_query": 9.835400225479143,
"max_relevant_docs_per_query": 18,
"unique_relevant_docs": 2835
},
"top_ranked_statistics": {
"num_top_ranked": 883511,
"min_top_ranked_per_query": 924,
"average_top_ranked_per_query": 996.0665163472379,
"max_top_ranked_per_query": 1068
}
},
"unseen_genus": {
"num_samples": 3565,
"num_queries": 730,
"num_documents": 2835,
"number_of_characters": 0,
"documents_text_statistics": null,
"documents_image_statistics": {
"min_image_width": 240,
"average_image_width": 1689.7855379188713,
"max_image_width": 2048,
"min_image_height": 143,
"average_image_height": 1347.004585537919,
"max_image_height": 2048,
"unique_images": 2813
},
"documents_audio_statistics": null,
"documents_video_statistics": null,
"queries_text_statistics": null,
"queries_image_statistics": null,
"queries_audio_statistics": {
"total_duration_seconds": 24517.693280116215,
"min_duration_seconds": 0.8950113378684807,
"average_duration_seconds": 33.58588120563865,
"max_duration_seconds": 615.0101979166667,
"unique_audios": 730,
"average_sampling_rate": 45114.31506849315,
"sampling_rates": {
"44100": 550,
"48000": 160,
"24000": 2,
"22050": 9,
"32000": 1,
"16000": 3,
"96000": 3,
"192000": 2
}
},
"queries_video_statistics": null,
"relevant_docs_statistics": {
"num_relevant_docs": 11129,
"min_relevant_docs_per_query": 1,
"average_relevant_docs_per_query": 15.245205479452055,
"max_relevant_docs_per_query": 95,
"unique_relevant_docs": 2835
},
"top_ranked_statistics": {
"num_top_ranked": 937519,
"min_top_ranked_per_query": 1053,
"average_top_ranked_per_query": 1284.272602739726,
"max_top_ranked_per_query": 1509
}
}
}
}
}
This dataset card was automatically generated using MTEB