Dynamic feature pyramid networks for object detection

Mingjian Zhu

doi:10.1117/12.3022812

28 March 2024 Dynamic feature pyramid networks for object detection

Mingjian Zhu

Proceedings Volume 13091, Fifteenth International Conference on Signal Processing Systems (ICSPS 2023); 130911N (2024) https://doi.org/10.1117/12.3022812
Event: Fifteenth International Conference on Signal Processing Systems (ICSPS 2023), 2023, Xi’an, China

Abstract

Feature pyramid network (FPN) is a critical component in modern object detection frameworks. The performance gain in most of the existing FPN variants is mainly attributed to the increase in computational burden. An attempt to enhance the FPN is enriching the spatial information by expanding the receptive fields, which is promising to largely improve the detection accuracy. In this paper, we first investigate how the expanding receptive fields affects the accuracy and computational costs of FPN. We explore a baseline model called inception FPN in which each lateral connection contains convolution filters with different kernel sizes. Moreover, we point out that not all objects need such a complicated calculation and propose a new dynamic FPN (DyFPN). The output features of DyFPN will be calculated by using the adaptively selected branch according to a dynamic gating operation. Therefore, the proposed method can provide a more efficient dynamic inference for achieving a better trade-off between accuracy and computational cost. Extensive experiments conducted on MS-COCO benchmark demonstrate that the proposed DyFPN significantly improves performance with the optimal allocation of computation resources. For instance, replacing inception FPN with DyFPN reduces about 40% of its FLOPs while maintaining a similar high performance.

(2024) Published by SPIE. Downloading of the abstract is permitted for personal use only.

Citation Download Citation

Mingjian Zhu "Dynamic feature pyramid networks for object detection", Proc. SPIE 13091, Fifteenth International Conference on Signal Processing Systems (ICSPS 2023), 130911N (28 March 2024); https://doi.org/10.1117/12.3022812

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available

Members: $17.00

Non-members: $21.00 ADD TO CART

PROCEEDINGS
9 PAGES

DOWNLOAD PAPER SAVE TO MY LIBRARY

GET CITATION

RIGHTS & PERMISSIONS

Get copyright permission Get copyright permission on Copyright Marketplace

KEYWORDS

Convolution

Object detection

Education and training

Tunable filters

Performance modeling

Ablation

Neural networks

Show All Keywords

Keywords/Phrases

Search In:

Publication Years