The deep convolutional neural network (CNN) has made remarkable progress in image classification. However, this network performs poorly and even cannot converge in many actual applications, where the training and test samples contain lots of noises. To solve the problems, this paper puts forward a network training strategy based on stochastic max pooling. Unlike the traditional max pooling, the proposed strategy first ranks all the values in each receptive field, and then selects a random value from the top-n values as the pooling result. Compared with common pooling methods, stochastic max pooling can limit the pooling selection to a larger value that represents the main information of the pooling area which reduces the chance of introducing noises into the network, and enhances the robustness of extracting noisy image features. Experimental results show that the CNN used stochastic max pooling Strategy can converge better than traditional CNN and classified noisy images much more accurately than traditional pooling methods.

In marine engineering and water conservancy projects, dredging often occurs due to silt accumulation, which can impede the long-term development of water ecosystems and water storage systems. To enhance dredging efficiency and the performance of dredging machines, a novel type of winch suction underwater dredging robot was designed. Computational fluid dynamics software was utilized to establish a fluid model of the robot’s winch suction dredging device and conduct model simulation experiments. The simulation test results were used to investigate the factors influencing dredging performance and their impact laws. Five key factors—namely, the reamer rotational speed, reamer arrangement angle, water flow rate, inlet pipe diameter, and outlet pipe diameter—were selected for consideration. By setting up various sets of factor levels, the significant influence of different factors on dredging efficiency was examined. Analysis of variance was employed to analyse the results of the orthogonal experimental design, leading to the identification of optimal factor levels and the establishment of an optimal scheme group. The results of the optimal scheme were verified, demonstrating a 13.049% increase in dredging efficiency and a 19.23% decrease in power consumption of the sludge pump compared to the initial experimental setup. The performance of the optimal program surpassed that of all the experimental designs and met the design requirements.

Object detection is essential for the perception systems of intelligent driving vehicles. RT-DETR has emerged as a prominent model. However, its direct application in intelligent driving vehicles still faces issues with the misdetection of occluded or small targets. To address these challenges, we propose a High-Precision Real-Time object detection algorithm (HPRT-DETR). We designed a Basic-iRMB-CGA (BIC) Block for a backbone network that efficiently extracts features and reduces the model’s parameters. We thus propose a Deformable Attention-based Intra-scale Feature Interaction (DAIFI) module by combining the Deformable Attention mechanism with the Intra-Scale Feature Interaction module. This enables the model to capture rich semantic features and enhance object detection accuracy in occlusion. The Local Feature Extraction Fusion (LFEF) block was created by integrating the local feature extraction module with the CNN-based Cross-scale Feature Fusion (CCFF) module. This integration expands the model’s receptive field and enhances feature extraction without adding learnable parameters or complex computations, effectively minimizing missed detections of small targets. Experiments on the KITTI dataset show that, compared to RT-DETR, HPRT-DETR improves mAP50 and FPS by 1.98% and 15.25%, respectively. Additionally, its generalization ability is assessed on the SODA 10M dataset, where HPRT-DETR outperforms RT-DETR in most evaluation metrics, confirming the model’s effectiveness.

Defect detection is crucial for quality control in industrial products. The defects in industrial products are typically subtle, leading to reduced accuracy in detection. Furthermore, industrial defect detection often necessitates high efficiency in order to meet operational demands. Deep learning-based algorithms for surface defect detection have been increasingly applied to industrial production processes. Among them, Swin-Transformer achieves remarkable success in many visual tasks. However, the computational burden imposed by numerous image tokens limits the application of Swin-Transformer. To enhance both the detection accuracy and efficiency, this paper proposes a linear attention mechanism based on pyramid pooling. It utilizes a more concise linear attention mechanism to reduce the computational load, thereby improving detection efficiency. Furthermore, it enhances global feature extraction capabilities through pyramid pooling, which improves the detection accuracy. Additionally, the incorporation of partial convolution into the model improves local feature extraction, further enhancing detection precision. Our model demonstrates satisfactory performance with minimal computational cost. It outperforms Swin-Transformer by 1.2% mAP and 52 FPS on the self-constructed SIM card slot defect dataset. When compared to the Swin-Transformer model on the public PKU-Market-PCB dataset, our model achieves an improvement of 1.7% mAP and 51 FPS. These results validate the universality of the proposed approach.

In recent years, multi-modal 3D object detection algorithms have experienced significant development. However, current algorithms primarily focus on designing overall fusion strategies for multi-modal features, neglecting finer-grained representations, which leads to a decline in the detection accuracy of small objects. To address this issue, this paper proposes the Instance-aware Fine-grained feature Enhancement Cross Modal Transformer (IFE-CMT) model. We designed an Instance feature Enhancement Module (IE-Module), which can accurately extract object features from multi-modal data and use them to enhance overall features while avoiding view transformations and maintaining low computational overhead. Additionally, we design a new point cloud branch network that effectively expands the network’s receptive field, enhancing the model’s semantic expression capabilities while preserving texture details of the objects. Experimental results on the nuScenes dataset demonstrate that compared to the CMT model, our proposed IFE-CMT model improves mAP and NDS by 2.1% and 0.8% on the validation set, respectively. On the test set, it improves mAP and NDS by 1.9% and a 0.7%. Notably, for small object categories such as bicycles and motorcycles, the mAP improved by 6.6% and 3.7%, respectively, significantly enhancing the detection accuracy of small objects.