Directly benefiting from the deep learning methods, object detection has witnessed a great performance boost in recent years. However, drone-view object detection remains challenging for two main reasons: (1) Objects of tiny-scale with more blurs w.r.t. ground-view objects offer less valuable information towards accurate and robust detection; (2) The unevenly distributed objects make the detection inefficient, especially for regions occupied by crowded objects. Confronting such challenges, we propose an end-to-end global-local self-adaptive network (GLSAN) in this paper. The key components in our GLSAN include a global-local detection network (GLDN), a simple yet efficient self-adaptive region selecting algorithm (SARSA), and a local super-resolution network (LSRN). We integrate a global-local fusion strategy into a progressive scale-varying network to perform more precise detection, where the local fine detector can adaptively refine the target’s bounding boxes detected by the global coarse detector via cropping the original images for higher-resolution detection. The SARSA can dynamically crop the crowded regions in the input images, which is unsupervised and can be easily plugged into the networks. Additionally, we train the LSRN to enlarge the cropped images, providing more detailed information for finer-scale feature extraction, helping the detector distinguish foreground and background more easily. The SARSA and LSRN also contribute to data augmentation towards network training, which makes the detector more robust. Extensive experiments and comprehensive evaluations on the VisDrone2019-DET benchmark dataset and UAVDT dataset demonstrate the effectiveness and adaptivity of our method. Towards an industrial application, our network is also applied to a DroneBolts dataset with proven advantages. Our source codes have been available at https://github.com/dengsutao/glsan .

Background: Dental plaque causes many common oral diseases (e.g., caries, gingivitis, and periodontitis). Therefore, plaque detection and control are extremely important for children’s oral health. The objectives of this study were to design a deep learning-based artificial intelligence (AI) model to detect plaque on primary teeth and to evaluate the diagnostic accuracy of the model. Methods: A conventional neural network (CNN) framework was adopted, and 886 intraoral photos of primary teeth were used for training. To validate clinical feasibility, 98 intraoral photos of primary teeth were assessed by the AI model. Additionally, tooth photos were acquired using a digital camera. One experienced pediatric dentist examined the photos and marked the regions containing plaque. Then, a plaque-disclosing agent was applied, and the areas with plaque were identified. After 1 week, the dentist drew the plaque area on the 98 photos taken by the digital camera again to evaluate the consistency of manual diagnosis. Additionally, 102 intraoral photos of primary teeth were marked to denote the plaque areas obtained by the AI model and the dentist to evaluate the diagnostic capacity of each approach based on lower-resolution photos. The mean intersection-over-union (MIoU) metric was employed to indicate detection accuracy. Results: The MIoU for detecting plaque on the tested tooth photos was 0.726 ± 0.165. The dentist’s MIoU was 0.695 ± 0.269 when first diagnosing the 98 photos taken by the digital camera and 0.689 ± 0.253 after 1 week. Compared to the dentist, the AI model demonstrated a higher MIoU (0.736 ± 0.174), and the results did not change after 1 week. When the dentist and the AI model assessed the 102 intraoral photos, the MIoU was 0.652 ± 0.195 for the dentist and 0.724 ± 0.159 for the model. The results of a paired t-test found no significant difference between the AI model and human specialist (P > .05) in diagnosing dental plaque on primary teeth. Conclusions: The AI model showed clinically acceptable performance in detecting dental plaque on primary teeth compared with an experienced pediatric dentist. This finding illustrates the potential of such AI technology to help improve pediatric oral health.

Thyroid ultrasonography is a widely-used clinical technique for nodule diagnosis in thyroid regions. However, it remains difficult to detect and recognize the nodules due to low contrast, high noise, and diverse appearance of nodules. In today's clinical practice, senior doctors could pinpoint nodules by analyzing global context features, local geometry structure, and intensity changes, which would require rich clinical experience accumulated from hundreds and thousands of nodule case studies. To alleviate doctors' tremendous labor in the diagnosis procedure, we advocate a machine learning approach to the detection and recognition tasks in this paper. In particular, we develop a multi-task cascade convolution neural network framework (MC-CNN) to exploit the context information of thyroid nodules. It may be noted that, our framework is built upon a large number of clinically-confirmed thyroid ultrasound images with accurate and detailed ground truth labels. Other key advantages of our framework result from a multi-task cascade architecture, two stages of carefully-designed deep convolution networks in order to detect and recognize thyroid nodules in a pyramidal fashion, and capturing various intrinsic features in a global-to-local way. Within our framework, the potential regions of interest after initial detection are further fed to the spatial pyramid augmented CNNs to embed multi-scale discriminative information for fine-grained thyroid recognition. Experimental results on $4309$ clinical ultrasound images have indicated that, our MC-CNN is accurate and effective for both thyroid nodules detection and recognition. For the correct diagnosis rate of malignant and benign thyroid nodules, its mAP performance can achieve up to 98.2% accuracy, which outperforms the common CNNs by $5$ % on average. In addition, we conduct rigorous user studies to confirm that our MC-CNN outperforms experienced doctors, yet only consuming roughly 2% (1/48) of doctors' examination time on average. Therefore, the accuracy and efficiency of our new method exhibit its great potential in clinical applications.

Dental plaque segmentation is crucial for maintaining oral health. However, accurately segmenting dental plaque in unconstrained environments can be challenging due to its low contrast and high variability in appearance. While existing transformer-based networks rely on attention mechanisms for each pixel, they do not take into account the relationships between neighboring pixels. Consequently, feature extraction is limited, making it difficult to achieve accurate segmentation of lowcontrast images. To address this issue, we propose a simple yet efficient cluster center transformer that improves dental plaque segmentation by clustering image pixels based on multiple levels of feature maps' intensity and texture information. By grouping similar pixels into regions, the proposed method enables the transformers to focus on the local contour and edge around the teeth regions, adapting to the low contrast and high variability of plaque appearance, leading to more accurate and efficient segmentation of dental plaque in dental images. Additionally, we designed Multiple Granularity Perceptions using a pyramid fusion mechanism to capture multiple scales of vision features, thereby enhancing the low-contrast vision features. The proposed method can benefit the dental diagnosis and treatment planning process by improving the accuracy and efficiency of dental plaque segmentation. Our proposed method achieved state-of-the-art results on the dental plaque dataset [1], with intersection over union (IoU) of 60.91% and pixel accuracy (PA) of 76.81%, all of which were the highest among all methods, demonstrating its effectiveness in plaque segmentation in unconstrained environments

Existing RGB-D salient object detection methods treat depth information as an independent component to complement RGB and widely follow the bistream parallel network architecture. To selectively fuse the CNN features extracted from both RGB and depth as a final result, the state-of-the-art (SOTA) bistream networks usually consist of two independent subbranches: one subbranch is used for RGB saliency, and the other aims for depth saliency. However, depth saliency is persistently inferior to the RGB saliency because the RGB component is intrinsically more informative than the depth component. The bistream architecture easily biases its subsequent fusion procedure to the RGB subbranch, leading to a performance bottleneck. In this paper, we propose a novel data-level recombination strategy to fuse RGB with D (depth) before deep feature extraction, where we cyclically convert the original 4-dimensional RGB-D into D GB, R D B and RG D . Then, a newly lightweight designed triple-stream network is applied over these novel formulated data to achieve an optimal channel-wise complementary fusion status between the RGB and D, achieving a new SOTA performance.

It is a challenging task to create realistic 3D avatars that accurately replicate individuals’ speech and unique talking styles for speech-driven facial animation. Existing techniques have made remarkable progress but still struggle to achieve lifelike mimicry. This article proposes “TalkingStyle”, a novel method to generate personalized talking avatars while retaining the talking style of the person. Our approach uses a set of audio and animation samples from an individual to create new facial animations that closely resemble their specific talking style, synchronized with speech. We disentangle the style codes from the motion patterns, allowing our method to associate a distinct identifier with each person. To manage each aspect effectively, we employ three separate encoders for style, speech, and motion, ensuring the preservation of the original style while maintaining consistent motion in our stylized talking avatars. Additionally, we propose a new style-conditioned transformer decoder, offering greater flexibility and control over the facial avatar styles. We comprehensively evaluate TalkingStyle through qualitative and quantitative assessments, as well as user studies demonstrating its superior realism and lip synchronization accuracy compared to current state-of-the-art methods.