University of Lincoln
Browse

Innovate Spatial-Temporal Attention Network (STAN) for Accurate 3D Mice Pose Estimation with a Single Monocular RGB Camera

Download (606.07 kB)
conference contribution
posted on 2025-02-04, 12:31 authored by Liyun GongLiyun Gong, Miao YuMiao Yu, Gautam Siddharth Kashyap, Sheldon McCallSheldon McCall, Mamatha ThotaMamatha Thota, Saeid Pourroostaei ArdakaniSaeid Pourroostaei Ardakani

  Precise 3D pose estimation of mice holds crucial  importance across various scientific domains. In this research,  we introduce an innovative model named the Spatial-Temporal  Attention Network (STAN), specifically designed for accurate  3D pose estimation of mice using a single monocular camera.  The STAN model leverages a sequence of extracted 2D skeleton  to predict the 3D pose of a mouse. Through the incorporation of  spatial and temporal attention modules, our STAN methodology  adeptly captures intricate spatial and temporal relationships  among key points, thereby enabling a comprehensive  representation of the dynamic movements inherent in a mouse's  behavior for precise 3D pose estimation. To assess the  effectiveness of our proposed method, extensive experimental  evaluations were undertaken. The results show the superior  performance of the STAN model when compared to other state of-the-art approaches within the realm of 3D mouse pose estimation.

History

School affiliated with

  • School of Computer Science (Research Outputs)

Publication Title

32nd European Signal Processing Conference (EUSIPCO 2024)

Publisher

IEEE

ISSN

2219-5491

eISSN

2076-1465

ISBN

979-8-3315-1977-3

eISBN

978-9-4645-9361-7

Date Accepted

2024-05-23

Date of Final Publication

2024-10-23

Event Name

32nd European Signal Processing Conference EUSIPCO 2024

Event Dates

26 - 30 August 2024

Open Access Status

  • Not Open Access

Publisher statement

© 2024 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.

Will your conference paper be published in proceedings?

  • Yes

Usage metrics

    University of Lincoln (Research Outputs)

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC