10K is Enough: An Ultra-Lightweight Binarized Network for Infrared Small-Target Detection
This work addresses model compression for infrared small-target detection on edge devices, but it is incremental as it builds on existing binarized neural network techniques.
The authors tackled the problem of deploying infrared small-target detection on edge devices by proposing BiisNet, a binarized network that integrates full-precision features to address precision loss, achieving strong competitiveness with state-of-the-art full-precision models.
The widespread deployment of Infrared Small-Target Detection (IRSTD) algorithms on edge devices necessitates the exploration of model compression techniques. Binarized neural networks (BNNs) are distinguished by their exceptional efficiency in model compression. However, the small size of infrared targets introduces stringent precision requirements for the IRSTD task, while the inherent precision loss during binarization presents a significant challenge. To address this, we propose the Binarized Infrared Small-Target Detection Network (BiisNet), which preserves the core operations of binarized convolutions while integrating full-precision features into the network's information flow. Specifically, we propose the Dot Binary Convolution, which retains fine-grained semantic information in feature maps while still leveraging the binarized convolution operations. In addition, we introduce a smooth and adaptive Dynamic Softsign function, which provides more comprehensive and progressively finer gradient during backpropagation, enhancing model stability and promoting an optimal weight distribution. Experimental results demonstrate that BiisNet not only significantly outperforms other binary architectures but also has strong competitiveness among state-of-the-art full-precision models.