This paper proposes a novel smart parking scheme for the parking lot. Automatic car detection is the core technology of the proposed scheme. However, new challenges arise in car detection in aerial views, such as a large number of tiny objects and complex backgrounds. In order to solve these issues, this paper proposes a car detection method based on multi-task cost-sensitive-convolutional neural network (MTCS-CNN). In the proposed network framework, multi-task partition layer which is composed of some sub-task selection units is first developed. The sub-task selection unit is constructed by introducing local mask and non-zero pooling, which can divide the complex detection task into many simple sub-tasks. To tackle the obtained sub-tasks, cost-sensitive sub-network is proposed based on faster R-CNN framework with the introduction of cost-sensitive loss. In the proposed Multi-task partition layer, the sub-task selection unit is used to capture the local map of the original aerial view image. In each local map, the scale and the number of objects are enlarged and decreased, respectively. Therefore, multi-task partition layer can divide a complex tiny objects detection task into many simple enlarged objects detection sub-tasks, which is helpful for performance improvement. In addition, the proposed cost-sensitive loss can effectively discount the effect of easy examples and focus attention on the hard examples, which may improve the detection performance on hard examples. Therefore, the model with incorporation of proposed cost-sensitive loss is robust to the complex background, further improving the performance. On our dataset, the proposed method obtained an mAP of 85.3%, outperformed state-of-the-art method.

Objective. Artificial intelligence‐powered screening systems of coronavirus disease 2019 (COVID‐19) are urgently demanding since the ongoing outbreak of SARS‐CoV‐2 worldwide. Chest CT or X‐ray is not sufficient to support the large‐scale screening of COVID‐19 because mildly‐infected patients do not have imaging features on these images. Therefore, it is imperative to exploit supplementary medical imaging strategies. Traditional Chinese medicine has played an essential role in the fight against COVID‐19. Methods. In this paper, we conduct two kinds of verification experiments based on a newly‐collected multi‐modality dataset, which consists of three types of modalities: tongue images, chest CT scans, and X‐ray images. First, we study a binary classification experiment on tongue images to verify the discriminative ability between COVID‐19 and non‐COVID‐19. Second, we design extensive multimodality experiments to validate whether introducing tongue image can improve the screening accuracy of COVID‐19 based on chest CT or X‐ray images. Results. Tongue image screening of COVID‐19 showed that the accuracy (ACC), sensitivity (SEN), specificity (SPEC), and Matthew correlation coefficient (MCC) of the improved AlexNet and Googlenet both reached 98.39%, 98.97%, 96.67%, and 99.11%. The fusion of chest CT and tongue images used a tandem multimodal classifier fusion strategy to achieve optimal classification, and the results and screening accuracy of COVID‐19 reached 98.98%, resulting in a significant improvement of 4.75% the highest accuracy in 375 years compared with the single‐modality model. The fusion of chest x‐rays and tongue images also had good classification accuracy. Conclusions. Both experimental results demonstrate that tongue image not only has an excellent discriminative ability for screening COVID‐19 but also can improve the screening accuracy based on chest CT or X‐rays. To the best of our knowledge, it is the first work that verifies the effectiveness of tongue image on screening COVID‐19. This paper provides a new perspective and a novel solution that contributes to large‐scale screening toward fast stopping the pandemic of COVID‐19.

Purpose: This study proposed a novel Local Reference Semantic Code (LRSC) network for automatic breast ultrasound image classification with few labeled data. Methods: In the proposed network, the local structure extractor is firstly developed to learn the local reference which describes common local characteristics of tumors. After that, a two-stage hierarchical encoder is developed to encode the local structures of lesion into the high-level semantic code. Based on the learned semantic code, the self-matching layer is proposed for the final classification. Results: In the experiment, the proposed method outperformed traditional classification methods and AUC (Area Under Curve), ACC (Accuracy), Sen (Sensitivity), Spec (Specificity), PPV (Positive Predictive Values), and NPV(Negative Predictive Values) are 0.9540, 0.9776, 0.9629, 0.93, 0.9774 and 0.9090, respectively. In addition, the proposed method also improved matching speed. Conclusions: LRSC-network is proposed for breast ultrasound images classification with few labeled data. In the proposed network, a two-stage hierarchical encoder is introduced to learn high-level semantic code. The learned code contains more effective high-level classification information and is simpler, leading to better generalization ability.