Recent deep models trained on large-scale RGB datasets lead to considerable achievements in visual detection tasks. However, the training examples are often limited for an infrared detection task, which may deteriorate the performance of deep detectors. In this paper, we propose a transfer approach, Source Model Guidance (SMG), where we leverage a high-capacity RGB detection model as the guidance to supervise the training process of an infrared detection network. In SMG, the foreground soft label generated from the RGB model is introduced as source knowledge to provide guidance for cross-domain transfer. Additionally, we design a Background Suppression Module in the infrared network to receive the knowledge and enhance the foreground features. SMG is easily plugged into any modern detection framework, and we show two explicit instantiations of it, SMG-C and SMG-Y, based on CenterNet and YOLOv3, respectively. Extensive experiments on different benchmarks show that both SMG-C and SMG-Y achieve remarkable performance even if the training set is scarce. Compared to advanced detectors on public FLIR, SMG-Y with 77.0% mAP outperforms others in accuracy, and SMG-C achieves real-time detection at a speed of 107 FPS. More importantly, SMG-Y trained on a quarter of the thermal dataset obtains 74.5% mAP, surpassing most state-of-the-art detectors with full FLIR as training data.

Robust detection of infrared slow-moving small targets is crucial in infrared search and tracking (IRST) applications such as infrared guidance and low-altitude security; however, existing methods easily cause missed detection and false alarms when detecting infrared small targets in complex low-altitude scenes. In this article, a new low-altitude slow-moving small target detection algorithm based on spatial-temporal features measure (STFM) is proposed. First, we construct a circular kernel to calculate the local grayscale difference (LGD) in a single image, which is essential to suppress low-frequency background and irregular edges in the spatial domain. Then, a short-term energy aggregation (SEA) mechanism with the accumulation of the moving target energy in multiple successive frames is proposed to enhance the dim target. Next, the spatial-temporal saliency map (STSM) is obtained by integrating the two above operations, and the candidate targets are segmented using an adaptive threshold mechanism from STSM. Finally, a long-term trajectory continuity (LTC) measurement is designed to confirm the real target and further eliminate false alarms. The SEA and LTC modules exploit the local inconsistency and the trajectory continuity of the moving small target in the temporal domain, respectively. Experimental results on six infrared image sequences containing different low-altitude scenes demonstrate the effectiveness of the proposed method, which performs better than the existing state-of-the-art methods.

Recent object detection networks suffer from performance degradation when training data and test data are distinct in image styles and content distributions. In this paper, we propose a domain adaptive method, Adversarial Mixing (AdvMix), where the label-rich source domain and unlabeled target domain are jointly trained by the adversarial feature alignment and a self-training strategy. To diminish the style gap, we design the Adversarial Gradient Reversal Layer (AdvGRL), containing a global-level domain discriminator to align the domain features by gradient reversal, and an adversarial weight mapping function to enhance the stability of domain-invariant features by hard example mining. To eliminate the content gap, we introduce a region mixing self-supervised training strategy where a region of the target image with the highest confidence is selected to merge with the source image, and the synthesis image is self-supervised by the consistency loss. To improve the reliability of self-training, we propose a strict confidence metric combining both object and bounding box uncertainty. Extensive experiments conducted on three benchmarks demonstrate that AdvMix achieves prominent performance in terms of detection accuracy, surpassing existing domain adaptive methods by nearly 5% mAP.