Search

The detection algorithm commonly misses obscured pedestrians in traffic scenes with a high pedestrian density because mutual occlusion among pedestrians reduces the prediction box score of the concealed pedestrians. The paper uses the YOLOv7 algorithm as the baseline and makes the following three improvements by investigating the variables influencing the detection method's performance: First, the backbone network of the YOLOv7 algorithm is replaced with the lightweight feature extraction network Mobilenetv3 since the pedestrian detection algorithm frequently needs to be deployed in driverless mobile, which requires a fast operating speed of the algorithm; second, a high-resolution feature pyramid structure is suggested for the issue of missed detection of hidden pedestrians, which upscales the feature maps generated from the feature pyramid to increase the resolution of the output feature maps and introduces shallow feature maps to strengthen the distinctions between adjacent sub-features to enhance the network's ability to extract features for the visible area of hidden pedestrians and small-sized pedestrians in order to produce deeper features with greater differentiation for pedestrians; and the third is to suggest a detection head based on an attention mechanism that is employed to lower the confidence level of target neighboring sub-features, lower the quantity of redundant detection boxes, and lower the following NMS computation. The mAP of the suggested approach in this work achieves 89.75%, which is 9.5 percentage points better than the YOLOv7 detection algorithm, according to experiments on the CrowdHuman pedestrian-intensive dataset. The algorithm proposed in this paper can considerably increase the detection performance of the detection algorithm, particularly for obscured pedestrians and small-sized pedestrians in the dataset, according to the experimental effect plots. [ABSTRACT FROM AUTHOR]

This paper proposes a multimodal fusion 3D target detection algorithm based on the attention mechanism to improve the performance of 3D target detection. The algorithm utilizes point cloud data and information from the camera. For image feature extraction, the ResNet50 + FPN architecture extracts features at four levels. Point cloud feature extraction employs the voxel method and FCN to extract point and voxel features. The fusion of image and point cloud features is achieved through regional point fusion and voxel fusion methods. After information fusion, the Coordinate and SimAM attention mechanisms extract fusion features at a deep level. The algorithm's performance is evaluated using the DAIR-V2X dataset. The results show that compared to the Part-A2 algorithm; the proposed algorithm improves the mAP value by 7.9% in the BEV view and 7.8% in the 3D view at IOU = 0.5 (cars) and IOU = 0.25 (pedestrians and cyclists). At IOU = 0.7 (cars) and IOU = 0.5 (pedestrians and cyclists), the mAP value of the SECOND algorithm is improved by 5.4% in the BEV view and 4.3% in the 3D view, compared to other comparison algorithms. [ABSTRACT FROM AUTHOR]

Infrared pedestrian detection has important theoretical research value and a wide range of application scenarios. Because of its special imaging method, infrared images can be used for pedestrian detection at night and in severe weather conditions. However, the lack of pedestrian feature information in infrared images and the small scale of pedestrian objects makes it difficult for detection networks to extract feature information and accurately detect small-scale pedestrians. To address these issues, this paper proposes an infrared pedestrian detection network based on YOLOv5, named IPD-Net. Firstly, an adaptive feature extraction module (AFEM) is designed in the backbone network section, in which a residual structure with stepwise selective kernel was included to enable the model to better extract feature information under different sizes of the receptive field. Secondly, a coordinate attention feature pyramid network (CA-FPN) is designed to enhance the deep feature map with location information through the coordinate attention module, so that the network gains better capability of object localization. Finally, shallow information is introduced into the feature fusion network to improve the detection accuracy of weak and small objects. Experimental results on the large infrared image dataset ZUT show that the mean Average Precision (mAP50) of our model is improved by 3.6% compared to that of YOLOv5s. In addition, IPD-Net shows various degrees of accuracy improvement compared to other excellent methods. [ABSTRACT FROM AUTHOR]