BNE-DETR: Nighttime Pedestrian Detection with Visible Light Sensors via Feature Enhancement and Multi-Scale Fusion

Fu Li, Yan Lu, Ming Zhao, Wangyu Wu · 2025

Read BNE-DETR: Nighttime Pedestrian Detection with Visible Light Sensors via Feature Enhancement and Multi-Scale Fusion by Fu Li, Yan Lu, Ming Zhao, Wangyu Wu (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-12788328.txt.

Description

Fu Li, Yan Lu, Ming Zhao, and Wangyu Wu proposed in 2025 a model called BNE-DETR to spot pedestrians at night with visible light cameras. At night, weak features, noise, and mixed target sizes make people hard to find. Their model builds on an existing real-time detector with three new parts.

One part sharpens feature extraction in low light with attention and gated units. A second part boosts weak detail and blends context across a scale. A third part fuses features across sizes with large-kernel and star-shaped blocks. On a night dataset, the model beat its base version on precision, recall, and average precision, while using fewer parameters and modest computing cost.

Catalog record

CREATORFu Li, Yan Lu, Ming Zhao, Wangyu Wu
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE81 KB
VERSION1.0
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC12788328

Browse the full library · About the Tower of Babel