About Wildlife2024
Wildlife2024 aims to provide researchers with a comprehensive and challenging platform for developing and evaluating wildlife visual analysis algorithms. Our goal is to advance the understanding of animal behavior, enhance wildlife conservation technologies, and promote the application of computer vision in ecological studies.
The dataset comprises video clips of diverse wildlife species captured across multiple environments, encompassing varying lighting conditions, occlusion scenarios, and animal behaviors.
We believe Wildlife2024 will serve as a pivotal resource for advancing research in related fields.
Our Wildlife2024 dataset comprises:
- Training dataset: Contains 1,450 sequences with over 5.1 million total frames.
- Testing dataset: Contains 110 sequences with over 100,000 total frames.
- Animal species: The dataset covers 8 animal classes: Birds, Fish, Amphibians, Mollusks, Mammals, Arthropods, Reptiles, and Coelenterates.
- Wild environments: The dataset encompasses 10 wild environments types: grasslands, forests, freshwater systems, oceans, deserts, wetlands, mountainous regions, polar regions, caves, and meadows.
- Challenge attributes: The dataset encompasses 13 challenge attributes: illumination variations (IV), out-of-plane rotations (OPR), in-plane rotations (IPR), deformation (DEF), fast motion (FM), scale variations (SV), camera motion (CM), out-of-view (OV), partial occlusion (POC), full occlusion (FOC), low resolution (LR), similar objects (SO), and motion blur (MB).
Animal species
Amphibians
Arthropods
Bird
Coelenterate
Fish
Mammals
Mollusc
Reptiles
Challenge attributes
IV
MB
LR
OPR
DEF
CM
OV
POC
SOB
FOC
IPR
SV
FM
Wildlife2024-test
White-Stork
Egret1
Egret2
Egret3
Bald-Eagle
Mottled Crow
Mottled Bunting
Polar bear1
Polar bear2
Northern Cardinal
Bat1
Bat2
Elephant1
Mantled guereza
Thrush1
Thrush2
Bee-fly1
Bee-fly2
Hummingbird1
Hummingbird2
Hummingbird3
Hummingbird4
hummingbird moth
Dove
Sea turtle
Seagull1
Seagull2
Dolphin
Black triggerfish
Brant
Red billed hornbill
Red Avadavat
Butterfly
Ring-billed Gull
Wasp
European Bee-eater
Domestic duck
Canary
peacock
Slug
Bluespotted-ribbontail-ray
Cheetah1
cheetah2
warbler
Deer moth
Green Peafowl
Sparrow
Millipede
Python
Elephant4
Sika deer
Honey bee
Kudu
dragonfly1
dragonfly2
Shark
Oriental Turtle Dove
Tit
Peacock Pansy
Muscovy duck
Pika
tree frog
Jellyfish1
Jellyfish2
Water buffalo1
Water buffalo2
Squirrel1
Squirrel2
Squirrel3
Squirrel4
Squirrel5
Pelican
Verditer Flycatcher
Snail
Magpie
Clownfish
Red panda
Bumblebee
Panda
Treecreeper
Duck
Eurasian Hobby
Swallow
Wild duck1
Wild duck2
Wild duck3
Eurasian Wryneck
Parrot
Eagle
Warthog
Mandarin Duck
Finch
Marsh Frog
Octopus
Crane fly
Giraffe1
Giraffe2
Long-tailed monkey
Echidna
Guineafowl
Robin
Spider1
Spider2
Spider3
Spotted Dove
Violet-headed Hummingbird
Brown bear
Elephant2
Elephant3
Hippopotamus
Dataset details
Our dataset Wildlife2024 includes:
- Training set: Contains 1,450 sequences, with a total of over 5.1M frames.
- Test set: Contains 110 sequences, with a total of over 100k frames.
Our benchmark dataset is designed for single object tracking (SOT), evaluating the ability to consistently track individual animal targets in complex scenarios. It provides standardized evaluation protocols and code.
Single Object Tracker Testing
| Tracker | Source | AUC | Norm.Prec. | Prec. |
|---|---|---|---|---|
| SMAT | WACV 2024 | 0.740 | 0.909 | 0.878 |
| MVT | BMVC 2023 | 0.717 | 0.898 | 0.851 |
| CTTrack | AAAI 2023 | 0.741 | 0.890 | 0.872 |
| Stark-GOT | ICCV 2021 | 0.718 | 0.874 | 0.840 |
| ETTrack | WACV 2023 | 0.703 | 0.876 | 0.820 |
| SiamBAN | CVPR 2020 | 0.698 | 0.881 | 0.836 |
| SiamRBO | CVPR 2022 | 0.690 | 0.866 | 0.822 |
| CNNInMo | IJCAI 2022 | 0.680 | 0.841 | 0.801 |
| SiamGAT | CVPR 2021 | 0.678 | 0.861 | 0.786 |
| SiamCAR | CVPR 2020 | 0.669 | 0.835 | 0.783 |
| TCTrack++ | PAMI 2023 | 0.668 | 0.848 | 0.791 |
| SiamTPN | WACV 2022 | 0.667 | 0.848 | 0.773 |
| SGDViT | ICRA 2023 | 0.641 | 0.827 | 0.768 |
Datasheet
Download Dataset
The Wildlife2024 dataset can be downloaded via the following link. We recommend using a download manager for a more stable downloading experience.
Current Version: v1.0
Before downloading, please ensure you have read and agreed to our Data License Agreement。
You can also access the data through our GitHub repository.
How to Cite
If you use the Wildlife2024 dataset or benchmark in your research, please cite our work as follows:
@inproceedings{WATS-DA,
title={Wild Animal Tracking with High Quality-SAM and Domain Adaptation},
author={Ganggang Huang and Mengyin Wang and Fasheng Wang and Fuming Sun and Haojie Li},
booktitle={Computer Vision for Animal Behavior Tracking and Modeling In conjunction with Computer Vision and Pattern Recognition 2024},
year={2024},
url={https://drive.google.com/file/d/1LDLmI9Xs2CkaOezgWVm3gNnUr21n2smR/view}
}
Contact
If you have any questions or suggestions, please contact us via email:hgg20210315@163.com
Or submit your questions on our GitHub Issues page.
License
The Wildlife2024 dataset (including both data and annotations) is released under CC BY-NC-SA 4.0
This means you may:
- Share — copy and redistribute the material in any medium or format
- Adapt — remix, transform, and build upon the material
Under the following terms:
- Attribution (BY) — You must give appropriate credit, provide a link to the license, and indicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use.
- NonCommercial (NC) — You may not use the material for commercial purposes.
- ShareAlike (SA) — If you remix, transform, or build upon the material, you must distribute your contributions under the same license as the original.
For full license details, please refer to the official license deed.
Associated code and tools may be released under different open-source licenses. Please check individual repository documentation for specific terms.