YOLSO: You Only Look Small Object

Small object detection is challenging and far from satisfactory. Most general object detectors suffer from two critical issues with small objects: (1) Feature extractor based on classification network cannot express the characteristics of small objects reasonably due to insufficient appearance information of targets and a large amount of background interference around them. (2) The detector requires a much higher location accuracy for small objects than for general objects. This paper proposes an effective and efficient small object detector YOLSO to address the above problems. For feature representation, we analyze the drawbacks in previous backbones and present a Half-Space Shortcut(HSSC) module to build a background-aware backbone. Furthermore, a coarse-to-fine Feature Pyramid Enhancement(FPE) module is introduced for layer-wise aggregation at a granular level to enhance the semantic discriminability. For loss function, we propose an exponential L1 loss to promote the convergence of regression, and a focal IOU loss to focus on prime samples with high classification confidence and high IOU. Both of them significantly improves the location accuracy of small objects. The proposed YOLSO sets state-of-the-art results on two typical small object datasets, MOCOD and VeDAI, at a speed of over 200 FPS. In the meantime, it also outperforms the baseline YOLOv3 by a wide margin on the common COCO dataset.     

Machine learning algorithm selection for windage alteration fault diagnosis of mine ventilation system

Machine learning algorithms have been widely used in mine fault diagnosis. The correct selection of the suitable algorithms is the key factor that affects the fault diagnosis. However, the impact of machine learning algorithms on the prediction performance of mine fault diagnosis models has not been fully evaluated. In this study, the windage alteration faults (WAFs) diagnosis models, which are based on K-nearest neighbor algorithm (KNN), multi-layer perceptron (MLP), support vector machine (SVM), and decision tree (DT), are constructed. Furthermore, the applicability of these four algorithms in the WAFs diagnosis is explored by a T-type ventilation network simulation experiment and the field empirical application research of Jinchuan No. 2 mine. The accuracy of the fault location diagnosis for the four models in both networks was 100%. In the simulation experiment, the mean absolute percentage error (MAPE) between the predicted values and the real values of the fault volume of the four models was 0.59%, 97.26%, 123.61%, and 8.78%, respectively. The MAPE for the field empirical application was 3.94%, 52.40%, 25.25%, and 7.15%, respectively. The results of the comprehensive evaluation of the fault location and fault volume diagnosis tests showed that the KNN model is the most suitable algorithm for the WAFs diagnosis, whereas the prediction performance of the DT model was the second-best. This study realizes the intelligent diagnosis of WAFs, and provides technical support for the realization of intelligent ventilation.     

Block size selection in rate-constrained geometry based point cloud compression

Multimedia Tools and Applications - In geometry-based point cloud compression, the geometry information is typically compressed using octree coding. In octree coding, the size of the blocks in the...     

基于推理注意力机制的二阶段网络图像修复

现有的图像修复方法在处理大面积缺失或高度纹理化的图像时，通常会产生扭曲的结构或与周围区域不一致的模糊纹理，无法重建合理的图像结构。为此，提出了一种基于推理注意力机制的二阶段网络图像修复方法。首先通过边缘生成网络生成合理的幻觉边缘信息，然后在图像补全网络完成图像的重建工作。为了进一步生成视觉效果更逼真的图像，提高图像修复的精确度，在图像补全网络采用推理注意力机制，有效控制了生成特征的不一致性，从而生成更有效的信息。所提方法在多个数据集上进行了实验验证，结果表明该图像修复方法的结构相似性指数达到了88.9%，峰值信噪比达到了25.56 dB，与现有的图像修复方法相比，该方法具有更高的图像修复精确度，生成的图像更逼真。     

On the design of secure primitives for real world applications

This editorial paper provides an overview of new advances on hardware security challenges and promising solutions for real world applications. It first presents the increasing hardware vulnerabilities with the emergence of diverse computing platforms. Then, it details the technical challenges, research opportunities, and potential solutions to cyber-attacks on -physical- devices.     

A novel solution of deep learning for endoscopic ultrasound image segmentation: enhanced computer aided diagnosis of gastrointestinal stromal tumor

Multimedia Tools and Applications - Gastrointestinal stromal tumor is one of the critical tumors that doctors do not suggest to get frequent endoscopy, so there is a need for a diagnosis system...     

Reducing web latency with coding-based fast multi-path loss recovery

Wireless Networks - TCP latency is critical to the performance of Web services. However, packet loss greatly impairs the TCP performance due to its poor loss recovery mechanisms. Recent work FUSO...     

Multiple clusterings of heterogeneous information networks

Machine Learning - Traditional clustering algorithms focus on a single clustering result; as such, they cannot explore potential diverse patterns of complex real world data. To deal with this...     

Looking ahead: Joint small group detection and tracking in crowd scenes

Small group detection and tracking in crowd scenes are basis for high level crowd analysis tasks. However, it suffers from the ambiguities in generating proper groups and in handling dynamic changes of group configurations. In this paper, we propose a novel delay decision-making based method for addressing the above problems, motivated by the idea that these ambiguities can be solved using rich temporal context. Specifically, given individual detections, small group hypotheses are generated. Then candidate group hypotheses across consecutive frames and their potential associations are built in a tree. By seeking for the best non-conflicting subset from the hypothesis tree, small groups are determined and simultaneously their trajectories are got. So this framework is called joint detection and tracking. This joint framework reduces the ambiguities in small group decision and tracking by looking ahead for several frames. However, it results in the unmanageable solution space because the number of track hypotheses grows exponentially over time. To solve this problem, effective pruning strategies are developed, which can keep the solution space manageable and also improve the credibility of small groups. Experiments on public datasets demonstrate the effectiveness of our method. The method achieves the state-of-the-art performance even in noisy crowd scenes.