Aerial Image Dehazing Network Compression: Towards Micro-UAV’S Real-Time Online Inferencing

Liu, Lulu; Zhang, Meng; Wang, Kaipeng; Zhang, Jiahui; Wang, Zichen

doi:10.1007/978-981-99-0923-0_44

Cited by 1 publication

(1 citation statement)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, there exists a trade-off between the efficiency and accuracy of lightweight dehazing approaches. Some approaches employ knowledge distillation [14], [15] or pruning techniques [16], which may sacrifice accuracy for efficiency. In contrast, other methods directly construct lightweight networks to address this issue.…”

Section: Related Workmentioning

confidence: 99%

LFD-Net: Lightweight Feature-Interaction Dehazing Network for Real-Time Remote Sensing Tasks

Jin,

Chen,

Tian

et al. 2023

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

Currently, remote sensing equipments are evolving towards intelligence and integration, incorporating edge computing techniques to enable real-time responses. One of the key challenges in enhancing downstream decision-making capabilities is the pre-processing step of image dehazing. Existing dehazing methods usually suffer from steep computational costs with densely connected residual modules, as well as difficulties in maintaining visual quality. To tackle these problems, we designed a lightweight Atmosphere Scattering Model (ASM) based network structure to extract, fuse and weight multiscale features. Our proposed LFD-Net demonstrates strong interpretability by exploiting the gated fusion module and attention mechanism to realize feature interactions between multi-level representations. The experimental results of LFD-Net on SOTS dataset reach an average Frequency Per Second (FPS) of 54.41, approximately 8 times faster than seven most popular methods with equivalent metrics. After image dehazing by LFD-Net, the performance of object detection is significantly improved. The mean Average Precision when IoU = 0.5

show abstract

Section: Related Workmentioning

confidence: 99%