Welcome to the page for the ML challenge “Automatic Estimation of Dog Age”
Challenge participants are kindly required to register for challenge participation by August 20th using this form. Only registered participants will be eligible for the prize.
Submissions are received (via email to firstname.lastname@example.org with subject DogAge Challenge Submission) until September 10, 2019. Multiple resubmissions are allowed until this deadline (up to 5 resubmissions by each participant) to improve the solutions.
The winners will be announced at the ICANN’19 conference. To receive the prize, the winners must attend the conference.
Automatic age estimation is a challenging problem attracting attention of the computer vision and pattern recognition communities due to its many practical applications. Articial neural networks, such as CNNs are a popular tool for tackling thiד problem, and several datasets which can be used for training models are available.
Despite the fact that dogs are the most well studied species in animal science, and that ageing processes in dogs are in many aspects similar to those of humans, the problem of age estimation for dogs has so far been overlooked.
The goal of this challenge is developing models that will accurately predict apparent dog age.
Read more about the challenge here.
This challenge is based on the DogAge dataset that has been carefully collected in a collaboration between animal and computer
scientists. It contains images of dogs, mapping them to one of the three classes young (0 -2 years), adult (2 – 5 years) or senior (> 6 years).
The dataset currently consists of two parts:
1. Expert data: contains 1373 images collected by animal scientists, sampling pet dogs, shelter dogs, laboratory dogs, working dogs and commercial kennel dogs. Their age and division into the three groups was carefully verified. The images are mostly high-quality portraits with the dog facing upwards forward.
2. Petfinder data: contains 26190 images collected using the APIs of Petfinder, a portal for pet adoption. The division of dogs into groups is not verfied, and there is a diversity of angles and distances of the photos. The raw data has been cleaned, removing photos with more than one dog, containing pets or large parts of humans, and low quality images.
The two parts of the dataset together with the testing set can be found here.
This challenge is organized by:
ICANN’19: International Conference on Artificial Neural Networks
Tech4Animals Lab, University of Haifa
ETU “LETI” St. Petersburg
School of Biology and Environmental Sciences, University of Salford
Solution Submission opens: May 1, 2019
Solution Submission closes: September 10, 2019
Winner Announcement: September 14-17, at the ICANN’19
The winner will be identified according to the criteria defined below and (s)he will receive a prize. It is expected that the winner as well as other participants will submit articles describing their methodological approaches for publication in a peer-reviewed journal (under discussion).
The criteria for success Three evaluation metrics described here will be used to identify the winner(s).
Solution format: The solution should be submitted as a csv file, whose name is the submitter’s surname (e.g., Zamansky.csv).
The file should contain two comma separated columns: image file name (dog1.jpg) and the predicted class: 0 – adult, 1 – senior, 2 – young.
The solutions should be sent via email to Anna Zamansky at email@example.com
Please do not hesitate to contact us in case of any questions.