Reduce Model Parameters Research Articles

DiaDet-R: A lightweight and accurate rotated detector for diatom detection in drowning diagnostics

In the field of forensic science, diatom test is widely utilized to determine the cause and location of death for bodies found in water. Diatom detection plays an important role in the diatom test. However, diatom detection faces challenges due to the imbalanced distribution of diatom features and the complexity of these features. To address these two challenges, we propose DiaDet-R, a diatom rotated detector for drowning diagnosis. Specifically, to tackle the first challenge of feature distribution imbalance, we perform customized data augmentation on our self-established diatom dataset to increase sample diversity. Further, we introduce the Low-level Enhancement Path Aggregation Feature Pyramid Network (LEPAFPN) to better extract tail features of intra-class features that exhibit a long-tail distribution. To address the second challenge of the feature complexity, we propose three strategies: (1) we devise a basic building block, Double Inception Depth-wise Convolution (Double IDC), to enhance the detector’s long-distance modeling capability; (2) we improve the horizontal detector to a rotated object detector to address the issues of overlapping and extensive invalid backgrounds caused by arbitrary angles and shapes of diatom targets; (3) to solve the feature misalignment issue in rotated object detection, we propose the Rotated Alignment Head (RAHead) to align target features. Extensive experimental results on our self-established diatom dataset show that DiaDet-R achieves mAP50 and mAP75 of 95.73% and 89.99%, respectively, with reduced model parameters and GFLOPs to 4.34M and 18.48G, validating that DiaDet-R can detect diatoms quickly and accurately, outperforming mainstream rotated detectors.

Optimizing tomato detection and counting in smart greenhouses: A lightweight YOLOv8 model incorporating high- and low-frequency feature transformer structures

ABSTRACT Tomato harvesting in intelligent greenhouses is crucial for reducing costs and optimizing management. Agricultural robots, as an automated solution, require advanced visual perception. This study proposes a tomato detection and counting algorithm based on YOLOv8 (TCAttn-YOLOv8). To handle small, occluded tomato targets in images, a new detection layer (NDL) is added to the Neck and Head decoupled structure, improving small object recognition. The ColBlock, a dual-branch structure leveraging Transformer advantages, enhances feature extraction and fusion, focusing on densely targeted regions and minimizing small object feature loss in complex backgrounds. C2fGhost and GhostConv are integrated into the Neck network to reduce model parameters and floating-point operations, improving feature expression. The WIoU (Wise-IoU) loss function is adopted to accelerate convergence and increase regression accuracy. Experimental results show that TCAttn-YOLOv8 achieves an mAP@0.5 of 96.31%, with an FPS of 95 and a parameter size of 2.7 M, outperforming seven lightweight YOLO algorithms. For automated tomato counting, the R2 between predicted and actual counts is 0.9282, indicating the algorithm’s suitability for replacing manual counting. This method effectively supports tomato detection and counting in intelligent greenhouses, offering valuable insights for robotic harvesting and yield estimation research.

Early Sweet Potato Plant Detection Method Based on YOLOv8s (ESPPD-YOLO): A Model for Early Sweet Potato Plant Detection in a Complex Field Environment

Traditional methods of pest control for sweet potatoes cause the waste of pesticides and land pollution, but the target detection algorithm based on deep learning can control the precise spraying of pesticides on sweet potato plants and prevent most pesticides from entering the land. Aiming at the problems of low detection accuracy of sweet potato plants and the complex of target detection models in natural environments, an improved algorithm based on YOLOv8s is proposed, which can accurately identify early sweet potato plants. First, this method uses an efficient network model to enhance the information flow in the channel, obtain more effective global features in the high-level semantic structure, and reduce model parameters and computational complexity. Then, cross-scale feature fusion and the general efficient aggregation architecture are used to further enhance the network feature extraction capability. Finally, the loss function is replaced with InnerFocaler-IoU (IFIoU) to improve the convergence speed and robustness of the model. Experimental results showed that the mAP0.5 and model size of the improved network reached 96.3% and 7.6 MB. Compared with the YOLOv8s baseline network, the number of parameters was reduced by 67.8%, the amount of computation was reduced by 53.1%, and the mAP0.5:0.95 increased by 3.5%. The improved algorithm has higher detection accuracy and a lower parameter and calculation amount. This method realizes the accurate detection of sweet potato plants in the natural environment and provides technical support and guidance for reducing pesticide waste and pesticide pollution.

YOFIR: High precise infrared object detection algorithm based on YOLO and FasterNet

Infrared images often suffer from issues such as blurriness and unclear object boundaries, and existing object detection algorithms are developed based on visible light images, which makes infrared object detection more challenging. Therefore, this paper proposes an infrared image enhancement method and an infrared object detection algorithm based on YOLO and FasterNet, named YOFIR. Specifically, we apply CHALE, Auto Gamma, histogram equalization, and bilateral filtering to process images individually, then fuse the results with different weights to address the poor imaging quality of infrared images. Moreover, we utilize the FasterNet network for multi-scale feature extraction to adapt to low-resolution infrared images. We also reduce model parameters through GSConv and propose a novel Efficient Multi-Scale Group Convolution module, EMSGC, which enhances feature fusion by processing feature maps from different channels, effectively improving detection accuracy. Finally, the DyHead Block is incorporated into the head to enhance the capability of infrared object detection. Experimental results on the HIT-UAV infrared remote sensing dataset show that the proposed algorithm achieves a 4% improvement in mAP0.5 compared to YOLOv8. Moreover, on the FLIR dataset, the algorithm shows a 1.6% improvement in mAP0.95 over YOLOv8, with significant advantages in terms of model parameters and FLOPs.

Three-Stage Global Channel Pruning for Resources-Limited Platform.

Deep neural networks (DNNs) have demonstrated remarkable performance in many fields, and deploying them on resource-limited devices has drawn more and more attention in industry and academia. Typically, there are great challenges for intelligent networked vehicles and drones to deploy object detection tasks due to the limited memory and computing power of embedded devices. To meet these challenges, hardware-friendly model compression approaches are required to reduce model parameters and computation. Three-stage global channel pruning, which involves sparsity training, channel pruning, and fine-tuning, is very popular in the field of model compression for its hardware-friendly structural pruning and ease of implementation. However, existing methods suffer from problems such as uneven sparsity, damage to the network structure, and reduced pruning ratio due to channel protection. To solve these issues, the present article makes the following significant contributions. First, we present an element-level heatmap-guided sparsity training method to achieve even sparsity, resulting in higher pruning ratio and improved performance. Second, we propose a global channel pruning method that fuses both global and local channel importance metrics to identify unimportant channels for pruning. Third, we present a channel replacement policy (CRP) to protect layers, ensuring that the pruning ratio can be guaranteed even under high pruning rate conditions. Evaluations show that our proposed method significantly outperforms the state-of-the-art (SOTA) methods in terms of pruning efficiency, making it more suitable for deployment on resource-limited devices.

DNE-YOLO: A method for apple fruit detection in Diverse Natural Environments

The apple industry, recognized as a pivotal sector in agriculture, increasingly emphasizes the mechanization and intelligent advancement of picking technology. This study innovatively applies a mist simulation algorithm to apple image generation, constructing a dataset of apple images under mixed sunny, cloudy, drizzling and foggy weather conditions called DNE-APPLE. It introduces a lightweight and efficient target detection network called DNE-YOLO. Building upon the YOLOv8 base model, DNE-YOLO incorporates the CBAM attention mechanism and CARAFE up-sampling operator to enhance the focus on apples. Additionally, it utilizes GSConv and the dynamic non-monotonic focusing mechanism loss function WIOU to reduce model parameters and decrease reliance on dataset quality. Extensive experimental results underscore the efficacy of the DNE-YOLO model, which achieves a detection accuracy (precision) of 90.7%, a recall of 88.9%, a mean accuracy (mAP50) of 94.3%, a computational complexity (GFLOPs) of 25.4G, and a parameter count of 10.46M across various environmentally diverse datasets. Compared to YOLOv8, it exhibits superior detection accuracy and robustness in sunny, drizzly, cloudy, and misty environments, making it especially suitable for practical applications such as apple picking for agricultural robots. The code for this model is open source at https://github.com/wuhaitao2178827/DNE-YOLO.

RCYOLO: an innovative surface defect detection model for magnetic blocks integrating RevCol and Yolov8s

ABSTRACT The surface defect detection of magnetic blocks faces challenges such as low accuracy, high missed detection rates, and weak anti-interference capabilities. Additionally, existing models are often large in size and require significant computational resources. To address these issues, this article integrated the Yolov8 framework and designed the RCYOLO model to improve detection performance and resource utilization. Firstly, we used RevCol as the backbone, and explored its optimal number of sub-networks to reduce model parameters and complexity, as well as improve model robustness. Secondly, we removed some deep structures from Yolov8s to simplify the model, making it easier to deploy. Afterwards, we developed a more efficient feature extraction module, C2f_SA, and replaced all existing modules with it to compensate for accuracy loss caused by simplification and improve overall detection performance. Finally, we optimized reg_max to explore the best minimum value for detecting defects in magnetic blocks to reduce unnecessary computational burden. Experimental results show that compared to Yolov8s, although RCYOLO’s mAP0.5 was reduced by 0.2%, it significantly reduced the model parameters, FLOPs, and model size by 51.64%, 30.31%, and 50.22%, respectively, achieving a better balance between accuracy and resources. Moreover, RCYOLO demonstrated good detection performance compared to other improved models.

MobilenetV2-RC: A lightweight network model for retinopathy classification in retinal OCT images

Abstract Retinopathy is an important ophthalmic disease that causes blindness in the elderly population. With the global old peoples increasing, the ophthalmic healthcare system is more and significant to pre-diagnosis. Optical Coherence Tomography (OCT) is considered the gold standard for ophthalmic treatment and diagnosis. OCT technologies and equipment continue to develop towards the intelligence and convenience for requirements of rapid diagnosis in the remote and poverty-stricken areas. We proposed an improved MobilenetV2 lightweight model for retinopathy classification (MobilenetV2-RC), which incorporates Spatial and Channel Reconstruction Convolution (SCConv) and the improved Convolutional Block Attention Module (CBAM) attention mechanism into the framework. Not only can it effectively limit feature redundancy to reduce model parameters, but also enhance the ability of feature representation to improve classification accuracy. The parameters of the proposed model are only 1.96 M with an overall accuracy of 98.96%, which is higher 3.32% than the original MobilenetV2. Compared with ResNet18, InceptionV3, and VGG16_BN, the overall accuracy is increased by 4.6%, 6.3%, and 3.9%, respectively. The test results of UCSD and Duke open-source datasets are more remarkable. Experimental results show that our proposed algorithm has strong reliability and generalization for the accurate classification of retinopathy, and a greater application prospect in the intelligent diagnosis of ophthalmology and mobile detection terminals.

A lightweight detection algorithm for tooth cracks in optical images

ObjectivesCracked tooth syndrome (CTS) is one of the major causes of tooth loss, presents the problem of early microcrack symptoms that are difficult to distinguish. This paper aims to investigate the practicality and feasibility of an improved object detection algorithm for automatically detecting cracks in dental optical images. MethodsA total of 286 teeth were obtained from Sun Yat-sen University and Guangdong University of Technology, and simulated cracks were generated using thermal expansion and contraction. Over 3000 images of cracked teeth were collected, including 360 real clinical images. To make the model more lightweight and better suited for deployment on embedded devices, this paper improves the YOLOv8 model for detecting tooth cracks through model pruning and backbone replacement. Additionally, the impact of image enhancement modules and coordinate attention modules on optimizing our model was analyzed. ResultsThrough experimental validation, we conclude that that model pruning reduction maintains performance better than replacing a lightweight backbone network on a tooth crack detection task. This approach achieved a reduction in parameters and GFLOPs by 16.8 % and 24.3 %, respectively, with minimal impact on performance. These results affirm the effectiveness of the proposed method in identifying and labeling tooth fractures. In addition, this paper demonstrated that the impact of image enhancement modules and coordinate attention mechanisms on YOLOv8's performance in the task of tooth crack detection was minimal. ConclusionsAn improved object detection algorithm has been proposed to reduce model parameters. This lightweight model is easier to deploy and holds potential for assisting dentists in identifying cracks on tooth surfaces.

Lightweight deep learning model for underwater waste segmentation based on sonar images

In recent years, the rapid accumulation of marine waste not only endangers the ecological environment but also causes seawater pollution. Traditional manual salvage methods often have low efficiency and pose safety risks to human operators, making automatic underwater waste recycling a mainstream approach. In this paper, we propose a lightweight multi-scale cross-level network for underwater waste segmentation based on sonar images that provides pixel-level location information and waste categories for autonomous underwater robots. In particular, we introduce hybrid perception and multi-scale attention modules to capture multi-scale contextual features and enhance high-level critical information, respectively. At the same time, we use sampling attention modules and cross-level interaction modules to achieve feature down-sampling and fuse detailed features and semantic features, respectively. Relevant experimental results indicate that our method outperforms other semantic segmentation models and achieves 74.66 % mIoU with only 0.68 M parameters. In particular, compared with the representative PIDNet Small model based on the convolutional neural network architecture, our method can improve the mIoU metric by 1.15 percentage points and can reduce model parameters by approximately 91 %. Compared with the representative SeaFormer T model based on the transformer architecture, our approach can improve the mIoU metric by 2.07 percentage points and can reduce model parameters by approximately 59 %. Our approach maintains a satisfactory balance between model parameters and segmentation performance. Our solution provides new insights into intelligent underwater waste recycling, which helps in promoting sustainable marine development.

Informatics-enhanced prediction of failure strength in skeletal muscle tissue

Skeletal muscles are susceptible to injury during daily activities and competitive sports. Reliable prediction of tissue failure strength is a major challenge due to the large uncertainty in the mechanical behavior of skeletal muscle tissue. The present study reveals a strong correlation between histological characteristics and skeletal muscle failure strength by means of mechanical and histological experiments. We propose a data-driven hybrid modeling approach that enables an effective integration of data science and informatics tools to capture tissue failure strength. Uncertainty in the tissue failure strength is propagated into the posterior information of reduced model parameters via the Bayesian inference framework and parameter space compression. A histological enhancement to the softening hyperelasticity model is made by linking a quantified tissue-scale histological characteristic and stiff model parameters using artificial neural networks. The model is applied to skeletal muscle tissue from different species and sites to assess its predictive capabilities for physiological differences. The results show that the approach can achieve reliable predictions of skeletal muscle tissue failure strength. The proposed approach can be extended to different scales to enrich the understanding of structure–property linkages for biomaterials.

Study on Nighttime Pedestrian Trajectory-Tracking from the Perspective of Driving Blind Spots

With the acceleration of urbanization and the growing demand for traffic safety, developing intelligent systems capable of accurately recognizing and tracking pedestrian trajectories at night or under low-light conditions has become a research focus in the field of transportation. This study aims to improve the accuracy and real-time performance of nighttime pedestrian-detection and -tracking. A method that integrates the multi-object detection algorithm YOLOP with the multi-object tracking algorithm DeepSORT is proposed. The improved YOLOP algorithm incorporates the C2f-faster structure in the Backbone and Neck sections, enhancing feature extraction capabilities. Additionally, a BiFormer attention mechanism is introduced to focus on the recognition of small-area features, the CARAFE module is added to improve shallow feature fusion, and the DyHead dynamic target-detection head is employed for comprehensive fusion. In terms of tracking, the ShuffleNetV2 lightweight module is integrated to reduce model parameters and network complexity. Experimental results demonstrate that the proposed FBCD-YOLOP model improves lane detection accuracy by 5.1%, increases the IoU metric by 0.8%, and enhances detection speed by 25 FPS compared to the baseline model. The accuracy of nighttime pedestrian-detection reached 89.6%, representing improvements of 1.3%, 0.9%, and 3.8% over the single-task YOLO v5, multi-task TDL-YOLO, and the original YOLOP models, respectively. These enhancements significantly improve the model’s detection performance in complex nighttime environments. The enhanced DeepSORT algorithm achieved an MOTA of 86.3% and an MOTP of 84.9%, with ID switch occurrences reduced to 5. Compared to the ByteTrack and StrongSORT algorithms, MOTA improved by 2.9% and 0.4%, respectively. Additionally, network parameters were reduced by 63.6%, significantly enhancing the real-time performance of nighttime pedestrian-detection and -tracking, making it highly suitable for deployment on intelligent edge computing surveillance platforms.

A portable real-time concrete bridge damage detection system

A drone detection system can achieve real-time inspection of concrete bridge damage. However, it is difficult to deploy existing detection algorithms on a portable system due to their high computational cost. Therefore, a portable damage detection algorithm called CD-YOLOv8 is designed. CD-YOLOv8 is operated on Jetson Xavier NX to detect drone-captured bridge images in real-time. The multi-scale feature adaptive fusion module is used to reduce model parameters and improve detection speed. The spatial context pyramid and deformable convolutional structure are introduced to improve the detection accuracy of tiny concrete bridge damage. Compared with the YOLOv8n, the mAP50 of CD-YOLOv8 is increased by 3.6 %, while the model parameters and FLOPs are decreased by 0.05 M and 0.7 G, respectively. CD-YOLOv8 is also tested on the Jetson Xavier NX with a detection speed of 44 FPS, which is 15.79 % faster than the YOLOv8n. Therefore, CD-YOLOv8 has high accuracy and excellent real-time performance.

On-Line Detection Method of Salted Egg Yolks with Impurities Based on Improved YOLOv7 Combined with DeepSORT.

Salted duck egg yolk, a key ingredient in various specialty foods in China, frequently contains broken eggshell fragments embedded in the yolk due to high-speed shell-breaking processes, which pose significant food safety risks. This paper presents an online detection method, YOLOv7-SEY-DeepSORT (salted egg yolk, SEY), designed to integrate an enhanced YOLOv7 with DeepSORT for real-time and accurate identification of salted egg yolks with impurities on production lines. The proposed method utilizes YOLOv7 as the core network, incorporating multiple Coordinate Attention (CA) modules in its Neck section to enhance the extraction of subtle eggshell impurities. To address the impact of imbalanced sample proportions on detection accuracy, the Focal-EIoU loss function is employed, adaptively adjusting bounding box loss values to ensure precise localization of yolks with impurities in images. The backbone network is replaced with the lightweight MobileOne neural network to reduce model parameters and improve real-time detection performance. DeepSORT is used for matching and tracking yolk targets across frames, accommodating rotational variations. Experimental results demonstrate that YOLOv7-SEY-DeepSORT achieves a mean average precision (mAP) of 0.931, reflecting a 0.53% improvement over the original YOLOv7. The method also shows enhanced tracking performance, with Multiple Object Tracking Accuracy (MOTA) and Multiple Object Tracking Precision (MOTP) scores of 87.9% and 73.8%, respectively, representing increases of 17.0% and 9.8% over SORT and 2.9% and 4.7% over Tracktor. Overall, the proposed method balances high detection accuracy with real-time performance, surpassing other mainstream object detection methods in comprehensive performance. Thus, it provides a robust solution for the rapid and accurate detection of defective salted egg yolks and offers a technical foundation and reference for future research on the automated and safe processing of egg products.

Small-target smoking detection algorithm based on improved YOLOv5

The use of general target detection algorithms for small-target detection is computationally costly and has a high missed detection rate. A lightweight small-target detection model based on YOLOv5 is proposed to address this issue.First, a maximum pooling layer is introduced to reduce the number of calculations. Second, Shuffle_Conv is designed to replace the ordinary convolutional layer to reduce model parameters. To further compress the model, the Add fusion method is used in the C3 module, while the GAC3 layer is designed with GhostNet. Finally, Mosaic_9 is introduced to improve the small-target detection without increasing the number of calculations. Compared with YOLOv5, computation and parameters of the improved model are reduced by 84.9% and 39.1%, respectively, and the accuracy is improved by 2%, which is more obvious than that of the original model.

Green pepper fruits counting based on improved DeepSort and optimized Yolov5s.

Green pepper yield estimation is crucial for establishing harvest and storage strategies. This paper proposes an automatic counting method for green pepper fruits based on object detection and multi-object tracking algorithm. Green pepper fruits have colors similar to leaves and are often occluded by each other, posing challenges for detection. Based on the YOLOv5s, the CS_YOLOv5s model is specifically designed for green pepper fruit detection. In the CS_YOLOv5s model, a Slim-Nick combined with GSConv structure is utilized in the Neck to reduce model parameters while enhancing detection speed. Additionally, the CBAM attention mechanism is integrated into the Neck to enhance the feature perception of green peppers at various locations and enhance the feature extraction capabilities of the model. According to the test results, the CS_YOLOv5s model of mAP, Precision and Recall, and Detection time of a single image are 98.96%, 95%, 97.3%, and 6.3 ms respectively. Compared to the YOLOv5s model, the Detection time of a single image is reduced by 34.4%, while Recall and mAP values are improved. Additionally, for green pepper fruit tracking, this paper combines appearance matching algorithms and track optimization algorithms from SportsTrack to optimize the DeepSort algorithm. Considering three different scenarios of tracking, the MOTA and MOTP are stable, but the ID switch is reduced by 29.41%. Based on the CS_YOLOv5s model, the counting performance before and after DeepSort optimization is compared. For green pepper counting in videos, the optimized DeepSort algorithm achieves ACP (Average Counting Precision), MAE (Mean Absolute Error), and RMSE (Root Mean Squared Error) values of 95.33%, 3.33, and 3.74, respectively. Compared to the original algorithm, ACP increases by 7.2%, while MAE and RMSE decrease by 6.67 and 6.94, respectively. Additionally, Based on the optimized DeepSort, the fruit counting results using YOLOv5s model and CS_YOLOv5s model were compared, and the results show that using the better object detector CS_YOLOv5s has better counting accuracy and robustness.

A Lightweight Rice Pest Detection Algorithm Using Improved Attention Mechanism and YOLOv8

Intelligent pest detection algorithms are capable of effectively detecting and recognizing agricultural pests, providing important recommendations for field pest control. However, existing recognition models have shortcomings such as poor accuracy or a large number of parameters. Therefore, this study proposes a lightweight and accurate rice pest detection algorithm based on improved YOLOv8. Firstly, a Multi-branch Convolutional Block Attention Module (M-CBAM) is constructed in the YOLOv8 network to enhance the feature extraction capability for pest targets, yielding better detection results. Secondly, the Minimum Points Distance Intersection over Union (MPDIoU) is introduced as a bounding box loss metric, enabling faster model convergence and improved detection results. Lastly, lightweight Ghost convolutional modules are utilized to significantly reduce model parameters while maintaining optimal detection performance. The experimental results demonstrate that the proposed method outperforms other detection models, with improvements observed in all evaluation metrics compared to the baseline model. On the test set, this method achieves a detection average precision of 95.8% and an F1-score of 94.6%, with a model parameter of 2.15 M, meeting the requirements of both accuracy and lightweightness. The efficacy of this approach is validated by the experimental findings, which provide specific solutions and technical references for intelligent pest detection.

Utilizing improved YOLOv8 based on SPD-BRSA-AFPN for ultrasonic phased array non-destructive testing

Non-destructive testing (NDT) is a technique for inspecting materials and their defects without causing damage to the tested components. Phased array ultrasonic testing (PAUT) has emerged as a hot topic in industrial NDT applications. Currently, the collection of ultrasound data is mostly automated, while the analysis of the data is still predominantly carried out manually. Manual analysis of scan image defects is inefficient and prone to instability, prompting the need for computer-based solutions. Deep learning-based object detection methods have shown promise in addressing such challenges recently. This approach typically demands a substantial amount of high-resolution, well-annotated training data, which is challenging to obtain in NDT. Consequently, it becomes difficult to detect low-resolution images and defects with varying positional sizes. This work proposes improvements based on the state-of-the-art YOLOv8 algorithm to enhance the accuracy and efficiency of defect detection in phased-array ultrasonic testing. The space-to-depth convolution (SPD-Conv) is imported to replace strided convolution, mitigating information loss during convolution operations and improving detection performance on low-resolution images. Additionally, this paper constructs and incorporates the bi-level routing and spatial attention module (BRSA) into the backbone, generating multiscale feature maps with richer details. In the neck section, the original structure is replaced by the asymptotic feature pyramid network (AFPN) to reduce model parameters and computational complexity. After testing on public datasets, in comparison to YOLOv8 (the baseline), this algorithm achieves high-quality detection of flat bottom holes (FBH) and aluminium blocks on the simulated dataset. More importantly, for the challenging-to-detect defect side-drilled holes (SDH), it achieves F1 scores (weighted average of precision and recall) of 82.50% and intersection over union (IOU) of 65.96%, representing an improvement of 17.56% and 0.43%. On the experimental dataset, the F1 score and IOU for FBH reach 75.68% (an increase of 9.01%) and 83.79%, respectively. Simultaneously, the proposed algorithm demonstrates robust performance in the presence of external noise, while maintaining exceptionally high computational efficiency and inference speed. These experimental results validate the high detection performance of the proposed intelligent defect detection algorithm for ultrasonic images, which contributes to the advancement of the smart industry.

YOLO-Chili: An Efficient Lightweight Network Model for Localization of Pepper Picking in Complex Environments

Currently, few deep models are applied to pepper-picking detection, and existing generalized neural networks face issues such as large model parameters, prolonged training times, and low accuracy. To address these challenges, this paper proposes the YOLO-chili target detection algorithm for chili pepper detection. Initially, the classical target detection algorithm YOLOv5 serves as the benchmark model. We introduce an adaptive spatial feature pyramid structure that combines the attention mechanism and the concept of multi-scale prediction to enhance the model’s detection capabilities for occluded and small target peppers. Subsequently, we incorporate a three-channel attention mechanism module to improve the algorithm’s long-distance recognition ability and reduce interference from redundant objects. Finally, we employ a quantized pruning method to reduce model parameters and achieve lightweight processing. Applying this method to our custom chili pepper dataset, we achieve an average precision (AP) value of 93.11% for chili pepper detection, with an accuracy rate of 93.51% and a recall rate of 92.55%. The experimental results demonstrate that YOLO-chili enables accurate and real-time pepper detection in complex orchard environments.

A Pruning and Distillation Based Compression Method for Sonar Image Detection Models

Accurate underwater target detection is crucial for the operation of autonomous underwater vehicles (AUVs), enhancing their environmental awareness and target search and rescue capabilities. Current deep learning-based detection models are typically large, requiring substantial storage and computational resources. However, the limited space on AUVs poses significant challenges for deploying these models on the embedded processors. Therefore, research on model compression is of great practical importance, aiming to reduce model parameters and computational load without significantly sacrificing accuracy. To address the challenge of deploying large detection models, this paper introduces an automated pruning method based on dependency graphs and successfully implements efficient pruning on the YOLOv7 model. To mitigate the accuracy degradation caused by extensive pruning, we design a hybrid distillation method that combines output-based and feature-based distillation techniques, thereby improving the detection accuracy of the pruned model. Finally, we deploy the compressed model on an embedded processor within an AUV to evaluate its performance. Multiple experiments confirm the effectiveness of our proposed method in practical applications.