AnimalWeb: A Large-Scale Hierarchical Dataset of Annotated Animal Faces

Khan, Muhammad Haris; McDonagh, John; Khan, Salman; Muhammad, Shafi; Arora, Aditya; Khan, Fahad Shahbaz; Shao, Ling; Tzimiropoulos, Georgios

doi:10.1109/cvpr42600.2020.00697

Cited by 34 publications

(14 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Guo et al (2020) recently developed a multispecies facial detector trained on 41 primate species and 4 carnivores resulting in detection accuracies of 0.91, 0.98, and 0.98 for golden snub-nosed monkeys Rhinopithecus roxellana, Tibetan macaques Macaca thibetana, and tigers Panthera tigris, respectively. In addition, Khan et al (2020) present AnimalWeb, an annotated dataset of animal faces for 334 species across 21 orders, which achieves a class-wise face detection mean average precision of 0.64. We only used images of bears where an individual identification was known, to avoid unintentionally training and testing a detector on images of a very low number of individuals, which could influence performance.…”

Section: Discussionmentioning

confidence: 99%

Multispecies facial detection for individual identification of wildlife: a case study across ursids

Clapham

Miller²,

Nguyen³

et al. 2022

Mamm Biol

View full text Add to dashboard Cite

To address biodiversity decline in the era of big data, replicable methods of data processing are needed. Automated methods of individual identification (ID) via computer vision are valuable in conservation research and wildlife management. Rapid and systematic methods of image processing and analysis are fundamental to an ever-growing need for effective conservation research and practice. Bears (ursids) are an interesting test system for examining computer vision techniques for wildlife, as they have variable facial morphology, variable presence of individual markings, and are challenging to research and monitor. We leveraged existing imagery of bears living under human care to develop a multispecies bear face detector, a critical part of individual ID pipelines. We compared its performance across species and on a pre-existing wild brown bear Ursus arctos dataset (BearID), to examine the robustness of convolutional neural networks trained on animals under human care. Using the multispecies bear face detector and retrained sub-applications of BearID, we prototyped an end-to-end individual ID pipeline for the declining Andean bear Tremarctos ornatus. Our multispecies face detector had an average precision of 0.91–1.00 across all eight bear species, was transferable to images of wild brown bears (AP = 0.93), and correctly identified individual Andean bears in 86% of test images. These preliminary results indicate that a multispecies-trained network can detect faces of a single species sufficiently to achieve high-performance individual classification, which could speed-up the transferability and application of automated individual ID to a wider range of taxa.

show abstract

Section: Discussionmentioning

confidence: 99%

Multispecies facial detection for individual identification of wildlife: a case study across ursids

Clapham

Miller²,

Nguyen³

et al. 2022

Mamm Biol

View full text Add to dashboard Cite

show abstract

“…Over 18K images and 20K annotations are collected from several popular 2D pose datasets, including COCO [30], 300W [45], AFLW [24], OneHand10K [55], DeepFasion2 [12], AP-10K [67], MacaquePose [25], Vinegar Fly [42], Desert Locust [13], CUB-200 [58], CarFusion [44], AnimalWeb [22], and Keypoint-5 [60]. Keypoint numbers are diverse across different categories, ranging from 8 to 68.…”

Section: Mulit-category Pose (Mp-100) Datasetmentioning

confidence: 99%

“…- [22] 21.9K 300VW [46] 218K ATRW [28] 9.5K AFLW [24] 25K Horse-10 [34] 8.1K 300W [45] 3.8K Animal-Pose [4] 6.1K WFLW [61] 10K MacaquePose [25] 13K InterHand2.6M [35] 2.6M AP-10K [67] 10K RHD [75] 41K CUB-200 [58] 12K CMU Panoptic [47] 15K PoseTrack18 [1] 23K OneHand10K [55] 10K AI Challenger [59] 300K FreiHand [76] 130K CrowdPose [27] 20K SynthHands [37] 63K OCHuman [72] 4.7K GANerated [36] 330K COCO [30] 200K ApolloCar3D [49] 70K MPII [2] 25K CarFusion [44] 63K JHMDB [16] 31K Keypoint-5 [60] 10K MHP [73] 25K DeepFashion2 [12] 80K Menpo [69] 9K…”

Section: Introductionmentioning

confidence: 99%

Pose for Everything: Towards Category-Agnostic Pose Estimation

Xu¹,

Jin²,

Wang³

et al. 2022

Preprint

View full text Add to dashboard Cite

Existing works on 2D pose estimation mainly focus on a certain category, e.g. human, animal, and vehicle. However, there are lots of application scenarios that require detecting the poses/keypoints of the unseen class of objects. In this paper, we introduce the task of Category-Agnostic Pose Estimation (CAPE), which aims to create a pose estimation model capable of detecting the pose of any class of object given only a few samples with keypoint definition. To achieve this goal, we formulate the pose estimation problem as a keypoint matching problem and design a novel CAPE framework, termed POse Matching Network (POMNet). A transformer-based Keypoint Interaction Module (KIM) is proposed to capture both the interactions among different keypoints and the relationship between the support and query images. We also introduce Multi-category Pose (MP-100) dataset, which is a 2D pose dataset of 100 object categories containing over 20K instances and is well-designed for developing CAPE algorithms. Experiments show that our method outperforms other baseline approaches by a large margin. Codes and data are available at https://github.com/luminxu/Pose-for-Everything.

show abstract

“…Biggs et al provide 20K 2D dog postures based on the Stanford Dog Dataset [11]. AnimalWeb provides 21K face images [12]. The Horse10 dataset comprises 8K horses [13].…”

Section: Related Workmentioning

confidence: 99%

AcinoSet: A 3D Pose Estimation Dataset and Baseline Models for Cheetahs in the Wild

Joska,

Clark,

Muramatsu

et al. 2021

Preprint

View full text Add to dashboard Cite

Animals are capable of extreme agility, yet understanding their complex dynamics, which have ecological, biomechanical and evolutionary implications, remains challenging. Being able to study this incredible agility will be critical for the development of next-generation autonomous legged robots. In particular, the cheetah (acinonyx jubatus) is supremely fast and maneuverable, yet quantifying its whole-body 3D kinematic data during locomotion in the wild remains a challenge, even with new deep learning-based methods. In this work we present an extensive dataset of free-running cheetahs in the wild, called AcinoSet, that contains 119, 490 frames of multi-view synchronized highspeed video footage, camera calibration files and 7, 588 humanannotated frames. We utilize markerless animal pose estimation to provide 2D keypoints. Then, we use three methods that serve as strong baselines for 3D pose estimation tool development: traditional sparse bundle adjustment, an Extended Kalman Filter, and a trajectory optimization-based method we call Full Trajectory Estimation. The resulting 3D trajectories, humanchecked 3D ground truth, and an interactive tool to inspect the data is also provided. We believe this dataset will be useful for a diverse range of fields such as ecology, neuroscience, robotics, biomechanics as well as computer vision. Code and data can be found at: https://github.com/African-Robotics-Unit/AcinoSet.

show abstract

AnimalWeb: A Large-Scale Hierarchical Dataset of Annotated Animal Faces

Cited by 34 publications

References 27 publications

Multispecies facial detection for individual identification of wildlife: a case study across ursids

Multispecies facial detection for individual identification of wildlife: a case study across ursids

Pose for Everything: Towards Category-Agnostic Pose Estimation

AcinoSet: A 3D Pose Estimation Dataset and Baseline Models for Cheetahs in the Wild

Contact Info

Product

Resources

About