Overview

Background

Collective animal behaviors are teeming with life and intricate behavioral patterns. The sweetfish, included in this dataset, offers a unique window into understanding animal navigation in water. For ethologists, ecologists, and mathematical and theoretical biologists, decoding these patterns is pivotal, but automatic tracking the navigation of sweetfish, especially when in schools, introduces a plethora of challenges.

Challenge

You can develop an automatic tracking model that can pinpoint the navigational patterns of sweetfish with the Sweetfish dataset including locations of 10 fishes. The dataset includes videos and bounding boxes of fishes. To analyze the behaviors recorded by videos, automatic detection and tracking of bounding boxes is necessary. We have train, development, and test datasets. By training your object detection and tracking algorithms using training dataset, you can estimate and submit the bounding box data in development and test datasets.

Prize

Travel Fee Support

At least one person from the top-3 team was encouraged to attend the award ceremony on-site or online. We will support the travel fee (within Japan) for the attendance of up to 1 member from each winning team, who lives in Japan.

Winners Obligations

As a condition to being awarded a Prize, a top-5 winner must fulfill the following obligations. The detailed instructions will be sent to top-5 winners after the final submission deadline.

• Submit your code so that we can check for cheating.
• Submit a short report paper that describes the award methodology.

Detail

Data

The Dataset for this challenge is SweatFish dataset (you can download from [Google drive Link] ). You can use all dataset listed below:

• Video (.mp4 in training, development, and test sets)
• Bounding boxes (training set only; .txt file in MOT format [https://motchallenge.net/instructions/])

Evaluation

The goal of this challenge is accurate tracking of 10 sweetfishes. Performance of your model will be evaluated based on HOTA (Higher Order Tracking Accuracy) score, which is a holistic and popular score in multi-object tracking (MOT). HOTA is designed to overcome many of the limitations of previous metrics. For example, you can see this web page for understanding.

Submission Format

To submit your results to this competition you must construct a submission zip file containing a single file named answer.txt. Sample submission files for Development Phase are available as follows:( submission.zip )

Rules for Fish Tracking Challenge 2024

A. Registration Required

Please submit a team registration form before you join this competition even if you are a solo challenger. This is required to get prizes. There are no limitations for the team size. However, if you are creating multiple teams within the same lab, be careful not to violate the rules of PRIVATE SHARING.

When you want to merge your team with another team, please contact the Fish Tracking Challenge Admin Team via email.

B. No Private Sharing Outside Teams

Privately sharing code or data outside of teams is not permitted. It’s OK to share code if made available to all participants on the forum.

C. External Dataset

You may use data other than the competition data to develop and test your submission. However, you will ensure the external data is publicly available to everyone without any cost.

Timeline

Cite

Tools