Robust Re-identification of Manta Rays from Natural Markings by Learning Pose Invariant Embeddings

Moskvyak, Olga; Maire, Frédéric; Armstrong, Asia O.; Dayoub, Feras; Baktashmotlagh, Mahsa

doi:10.1109/dicta52665.2021.9647359

Cited by 14 publications

(17 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Wildbook uses PIE and Hotspotter metricbased algorithms to re-identify animals by pattern. PIE (Moskvyak, Maire, Dayoub, Armstrong, & Baktashmotlagh, 2021) is a deep learning-based method for matching of individuals invariantly to the pose. It receives shape embedding and pose embedding separately and normalizes the shape to match the individual regardless of the specific pose.…”

Section: Related Work 21 Animal Re-identificationmentioning

confidence: 99%

“…It receives shape embedding and pose embedding separately and normalizes the shape to match the individual regardless of the specific pose. PIE was originally developed for manta rays (Moskvyak, Maire, Dayoub, Armstrong, & Baktashmotlagh, 2021), but in Wildbook it is also used for humpback whale flukes, orcas, and right whales. HotSpotter (Crall et al, 2013) is a SIFT-based (D.G.…”

Section: Related Work 21 Animal Re-identificationmentioning

confidence: 99%

“…A variety of methods for animal reidentification exist that utilize distinct characteristics in fur, feather and skin patterns (T. Berger-Wolf et al, 2015;Crall, Stewart, Berger-Wolf, Rubenstein, & Sundaresan, 2013;Li, Li, Tang, Qian, & Lin, 2020;Moskvyak, Maire, Dayoub, Armstrong, & Baktashmotlagh, 2021), and methods originally developed for human face re-identification have been successfully applied to animals (Agarwal et al, 2019;Crouse et al, 2017;Deb et al, 2018). Visual animal re-identification can be formulated as a task of finding a match for the given query image from a database of known individuals, which is equivalent to a content-based image retrieval (CBIR) problem (Smeulders, Worring, Santini, Gupta, & Jain, 2000) where an image is searched from a database based on the image content.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

NORPPA: NOvel Ringed seal re-identification by Pelage Pattern Aggregation

Nepovinnykh¹,

Chelak²,

Eerola³

et al. 2022

Preprint

View full text Add to dashboard Cite

We propose a method for Saimaa ringed seal (Pusa hispida saimensis) re-identification. Access to large image volumes through camera trapping and crowdsourcing provides novel possibilities for animal monitoring and conservation and calls for automatic methods for analysis, in particular, when re-identifying individual animals from the images. The proposed method NOvel Ringed seal re-identification by Pelage Pattern Aggregation (NORPPA) utilizes the permanent and unique pelage pattern of Saimaa ringed seals and content-based image retrieval techniques. First, the query image is preprocessed, and each seal instance is segmented. Next, the seal's pelage pattern is extracted using a U-net encoder-decoder based method. Then, CNN-based affine invariant features are embedded and aggregated into Fisher Vectors. Finally, the cosine distance between the Fisher Vectors is used to find the best match from a database of known individuals. We perform extensive experiments of various modifications of the method on a new challenging Saimaa ringed seals re-identification dataset. The proposed method is shown to produce the best re-identification accuracy on our dataset in comparisons with alternative approaches.

show abstract

Section: Related Work 21 Animal Re-identificationmentioning

confidence: 99%

Section: Related Work 21 Animal Re-identificationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

NORPPA: NOvel Ringed seal re-identification by Pelage Pattern Aggregation

Nepovinnykh¹,

Chelak²,

Eerola³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…The muzzle patterns are obtained manually, providing consistent data that simplifies the re-identification. Convolutional neural network (CNN) approaches for animal re-identification by using natural body markings have been applied to various animals including manta rays [ 14 ], Amur tigers ( Panthera tigris altaica ) [ 15 , 16 , 17 ], zebras ( Equus grevyi ), and giraffes ( Giraffa ) [ 18 ]. Some species, such as bottlenose dolphins ( Tursiops truncatus ) or African savanna elephants ( Loxodonta africana ) can be identified based on the shape of their body parts, usually their tail or fins, or an ear in case of an elephant.…”

Section: Related Workmentioning

confidence: 99%

SealID: Saimaa Ringed Seal Re-Identification Dataset

Nepovinnykh

Eerola

Biard

et al. 2022

Sensors

View full text Add to dashboard Cite

Wildlife camera traps and crowd-sourced image material provide novel possibilities to monitor endangered animal species. The massive data volumes call for automatic methods to solve various tasks related to population monitoring, such as the re-identification of individual animals. The Saimaa ringed seal (Pusa hispida saimensis) is an endangered subspecies only found in Lake Saimaa, Finland, and is one of the few existing freshwater seal species. Ringed seals have permanent pelage patterns that are unique to each individual and that can be used for the identification of individuals. A large variation in poses, further exacerbated by the deformable nature of seals, together with varying appearance and low contrast between the ring pattern and the rest of the pelage makes the Saimaa ringed seal re-identification task very challenging, providing a good benchmark by which to evaluate state-of-the-art re-identification methods. Therefore, we make our Saimaa ringed seal image (SealID) dataset (N = 57) publicly available for research purposes. In this paper, the dataset is described, the evaluation protocol for re-identification methods is proposed, and the results for two baseline methods—HotSpotter and NORPPA—are provided. The SealID dataset has been made publicly available.

show abstract

“…The muzzle patterns are obtained manually, providing consistent data that simplifies the re-identification. CNN approaches for animal re-identification using natural body markings have been applied to various animals including manta rays [24], Amur tigers (Panthera tigris altaica) [18,17,19], zebras (Equus grevyi ), and giraffes (Giraffa) [33]. Some species, such as bottlenose dolphins (Tursiops truncatus) or African savanna elephants (Loxodonta africana) can be identified based on the shape of their body parts, usually their tail or fins, or an ear in case of an elephant.…”

Section: Animal Re-identificationmentioning

confidence: 99%

SealID: Saimaa ringed seal re-identification dataset

Nepovinnykh¹,

Eerola²,

Biard³

et al. 2022

Preprint

View full text Add to dashboard Cite

Wildlife camera traps and crowd-sourced image material provide novel possibilities to monitor endangered animal species. However, massive image volumes that these methods produce are overwhelming for researchers to go through manually which calls for automatic systems to perform the analysis. The analysis task that has gained the most attention is the re-identification of individuals, as it allows, for example, to study animal migration or to estimate the population size. The Saimaa ringed seal (Pusa hispida saimensis) is an endangered subspecies only found in the Lake Saimaa, Finland, and is one of the few existing freshwater seal species. Ringed seals have permanent pelage patterns that are unique to each individual which can be used for the identification of individuals. Large variation in poses further exacerbated by the deformable nature of seals together with varying appearance and low contrast between the ring pattern and the rest of the pelage makes the Saimaa ringed seal re-identification task very challenging, providing a good benchmark to evaluate state-of-the-art re-identification methods. Therefore, we make our Saimaa ringed seal image (SealID) dataset (N=57) publicly available for research purposes. In this paper, the dataset is described, the evaluation protocol for re-identification methods is proposed, and the results for two baseline methods HotSpotter and NORPPA are provided. The SealID dataset has been made publicly available at

show abstract

Robust Re-identification of Manta Rays from Natural Markings by Learning Pose Invariant Embeddings

Cited by 14 publications

References 21 publications

NORPPA: NOvel Ringed seal re-identification by Pelage Pattern Aggregation

NORPPA: NOvel Ringed seal re-identification by Pelage Pattern Aggregation

SealID: Saimaa Ringed Seal Re-Identification Dataset

SealID: Saimaa ringed seal re-identification dataset

Contact Info

Product

Resources

About