Authors: Qianchen Mao, Qiang Li, Bingshu Wang, Yongjun Zhang, Tao Dai, C.L. Philip Chen

Published on: February 08, 2024

Impact Score: 8.22

Arxiv code: Arxiv:2402.05410

Summary

• What is new: SpirDet introduces a unique dual-branch sparse decoder and a lightweight DO-RepEncoder for efficient infrared small target detection.
• Why this is important: Existing methods for detecting infrared small targets are computationally expensive and inefficient due to the preservation of high-resolution features.
• What the research proposes: SpirDet employs a dual-branch sparse decoder to identify target locations with minimal area coverage and fine-tunes these locations, alongside a DO-RepEncoder to reduce memory and speed up inference.
• Results: On the IRSTD-1K dataset, SpirDet achieves a 4.7 increase in MIoU and a 7 times faster FPS than the best previous models.

Technical Details

Technological frameworks used: nan

Models used: Dual-branch sparse decoder, DO-RepEncoder

Data used: IRSTD-1K dataset

Potential Impact

Security and surveillance industries, autonomous vehicle manufacturers, and any sector reliant on infrared imaging for small target detection could significantly benefit.

Want to implement this idea in a business?

We have generated a startup concept here: InfraEye.