Overview

Abstract

Multi-target multi-camera tracking is a crucial task that involves identifying and tracking individuals over time using video streams from multiple cameras. This task has practical applications in various fields, such as visual surveillance, crowd behavior analysis, and anomaly detection. However, due to the difficulty and cost of collecting and labeling data, existing datasets for this task are either synthetically generated or artificially constructed within a controlled camera network setting, which limits their ability to model real-world dynamics and generalize to diverse camera configurations. To address this issue, we present MTMMC, a real-world, large-scale dataset that includes long video sequences captured by 16 multi-modal cameras in two different environments - campus and factory - across various time, weather, and season conditions. This dataset provides a challenging test bed for studying multi-camera tracking under diverse real-world complexities and includes an additional input modality of spatially aligned and temporally synchronized RGB and thermal cameras, which enhances the accuracy of multi-camera tracking. MTMMC is a super-set of existing datasets, benefiting independent fields such as person detection, re-identification, and multiple object tracking. We provide baselines and new learning setups on this dataset and set the reference scores for future studies. The datasets, models, and test server will be made publicly available.

 DOWNLOAD

Data

We provide the video frames and annotations:

• Frames: JPG format

• Annotations:  COCO style JSON format

Dataset Access Request Instructions

You can request access to the MTMMC dataset by contacting us at gritycda@gmail.com.

Submission Guidelines

Failure to follow any of the following submission guidelines will result in your request not being processed. Please ensure you strictly adhere to all instructions when submitting your request.

1. Include the Dataset Terms and Conditions:

   • Attach a completed copy of the dataset terms and conditions form. [Link to form]

   • Download and fill out the MTMMC User Agreement.docx file.

   • Ensure all user information fields are completed.

   • Provide a detailed description of your intended use of the dataset (this will be used to evaluate your request, especially regarding ethical considerations).

   • Check the agreement box and insert your handwritten signature.

   • Delete any blue instructional text from the form before submission.

   • Convert the document to PDF format before sending.

   • File naming convention: Use the following format when saving your file: {Application Date}_{Affiliation}_{Your Name} (e.g., 20240623_KAIST_Jaden Kim).

2. Email Submission Requirements:

   • Use the subject line format: [MTMMC benchmark request] {Your Name}, {Affiliation}.

   • In the email body, include:

     • The Google email address to which the MTMMC benchmark drive will be shared.

     • A brief explanation of your intended use of the MTMMC benchmark.

Important Notice

Please note that unauthorized sharing or transfer of the dataset outside the methods outlined in the approval process is a violation of international law and may result in legal action. Therefore, all users must submit the signed agreement before accessing the dataset.

Revised: 10 / 16 / 2024

Acknowledgment

We gratefully acknowledge that MTMMC was built as part of the AI Training Data Construction Project 2021 hosted by the Ministry of Science and ICT (MSIT) and supported by the National Information Society Agency (NIA) of South Korea.