MammalNet: A Large-scale Video Benchmark for Mammal Recognition and Behavior Understanding (CVPR 2023)

Jun Chen, Ming Hu, Darren J. Coker, blair Costelloe, Michael L. Berumen, Sara Beery, Anna Rohrbach and Mohamed Elhoseiny.

Introduction

MammalNet is built around a biological mammal taxonomy spanning 17 orders, 69 families and 173 mammal categories, and includes 12 common high-level mammal behaviors (e.g. hunt, groom). We establish three benchmarks on MammalNet: standard animal and behavior recognition, compositional low-shot animal and behavior recognition, and behavior detection.

Dataset Download

To download the required files, follow these steps:

1. Trimmed Video Files:

   wget https://mammalnet.s3.amazonaws.com/trimmed_video.tar.gz

2. Annotation Files:

   wget https://mammalnet.s3.amazonaws.com/annotation.tar

3. Full-length Video Files:

   wget https://mammalnet.s3.amazonaws.com/full_video.tar.gz

Running our baseline codes

Recognition

Please refer recognition page to prepare the dataset and run code

Detection

Please refer detection page to prepare the dataset and run code

@InProceedings{Chen_2023_CVPR,
    author    = {Chen, Jun and Hu, Ming and Coker, Darren J. and Berumen, Michael L. and Costelloe, Blair and Beery, Sara and Rohrbach, Anna and Elhoseiny, Mohamed},
    title     = {MammalNet: A Large-Scale Video Benchmark for Mammal Recognition and Behavior Understanding},
    booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
    month     = {June},
    year      = {2023},
    pages     = {13052-13061}
}