Conventional Object Detection Research Articles

EMPViT: Efficient multi-path vision transformer for security risks detection in power distribution network

To maintain the safe operation of power distribution network (PDN) equipment, it is important to accurately and promptly identify security risks. However, conventional drone-based object detection methods face challenges due to noise and similarity features in risk targets, as well as limited computing resources of unmanned aerial vehicles (UAVs). To address these challenges, an efficient embedding-based multi-path fusion architecture is proposed. This architecture uses a re-parameterized depthwise block to embed local context information at different scales, enhancing the extraction of tiny features while preserving inference speed. Additionally, a coordinated self-attention module is proposed to reduce computational complexity while maintaining the performance of global information. By fusing fine and coarse feature representations without requiring a lot of computation, this module efficiently learns from both local and global features from images. The goal is to create an efficient multi-path vision transformer (EMPViT) architecture that achieves a balance between accuracy and efficiency. The proposed EMPViT has been evaluated on two different drone image dataset, demonstrating better performance compared to other architectures. Specifically, the EMPViT-S improves the detection mAP by 1.2%, and the inference speed is improved to 1.24 times on average on Drone-PDN dataset. It has achieved the same performance improvement on VisDrone-DET2019 dataset, gaining detection performance by 1.3% and 1.2 times acceleration on average.

Research on Low-Light Environment Object Detection Algorithm Based on YOLO_GD

In low-light environments, the presence of numerous small, dense, and occluded objects challenges the effectiveness of conventional object detection methods, failing to achieve desirable results. To address this, this paper proposes an efficient object detection network, YOLO_GD, which is designed for precise detection of targets in low-light scenarios. This algorithm, based on the foundational framework of YOLOv5s, implements a cross-layer feature fusion method founded on an information gathering and distribution mechanism. This method mitigates the issue of information loss during inter-layer feature exchange and, building on this, constructs a Bi-level routing spatial attention module to reduce computational redundancy caused by the self-attention mechanism, thereby enhancing the model’s detection accuracy for small objects. Furthermore, through the introduction of a novel deformable convolution, a cross-stage local feature fusion module is established, enabling the model to capture the complex features of input data more accurately and improve detection precision for dense objects. Lastly, the introduction of a probabilistic distance metric in the bounding box regression loss function enhances the network model’s generalization capability, further increasing detection accuracy in occluded scenarios. Experimental results on the ExDark dataset demonstrate that compared to YOLOv5, there is a 5.97% improvement in mean average precision (mAP), effectively enhancing object detection performance in low-light conditions.

LUD-YOLO: A novel lightweight object detection network for unmanned aerial vehicle

Autonomous execution of tasks by unmanned aerial vehicles (UAVs) relies heavily on object detection. However, object detection in most images presents challenges such as complex backgrounds, small targets, and obstructions. Additionally, the limited computing speed and memory of the UAV processor affects the accuracy of conventional object detection algorithms. This paper proposes LUD-You Only Look Once (YOLO), a small and lightweight object detection algorithm for UAVs based on YOLOv8. The proposed algorithm introduces a new multiscale feature fusion mode that solves the degradation in feature propagation and interaction through the introduction of upsampling in the feature pyramid network and the progressive feature pyramid network. The application of the dynamic sparse attention mechanism in the Cf2 module achieves flexible computing allocation and content awareness. Furthermore, the proposed model is optimized to be sparse and lightweight, making it possible to deploy on UAV edge devices. Finally, the effectiveness and superiority of LUD-YOLO were verified on the VisDrone2019 and UAVDT datasets. The results of ablation and comparison experiments show that compared with the original algorithm, LUDY-N and LUDY-S have shown excellent performance in various evaluation indexes, indicating that the proposed improvement strategies make the model have better robustness and generalization. Moreover, compared with multiple other popular competitors, the proposed improvement strategies enable LUD-YOLO to have the best overall performance, providing an effective solution for UAVs object detection while balancing model size and detection accuracy.

Improved 3D Object Detection Based on PointPillars

Despite the recent advancements in 3D object detection, the conventional 3D point cloud object detection algorithms have been found to exhibit limited accuracy for the detection of small objects. To address the challenge of poor detection of small-scale objects, this paper adopts the PointPillars algorithm as the baseline model and proposes a two-stage 3D target detection approach. As a cutting-edge solution, point cloud processing is performed using Transformer models. Additionally, a redefined attention mechanism is introduced to further enhance the detection capabilities of the algorithm. In the first stage, the algorithm uses PointPillars as the baseline model. The central concept of this algorithm is to transform the point cloud space into equal-sized columns. During the feature extraction stage, when the features from all cylinders are transformed into pseudo-images, the proposed algorithm incorporates attention mechanisms adapted from the Squeeze-and-Excitation (SE) method to emphasize and suppress feature information. Furthermore, the 2D convolution of the traditional backbone network is replaced by dynamic convolution. Concurrently, the addition of the attention mechanism further improves the feature representation ability of the network. In the second phase, the candidate frames generated in the first phase are refined using a Transformer-based approach. The proposed algorithm applies channel weighting in the decoder to enhance channel information, leading to improved detection accuracy and reduced false detections. The encoder constructs the initial point features from the candidate frames for encoding. Meanwhile, the decoder applies channel weighting to enhance the channel information, thereby improving the detection accuracy and reducing false detections. In the KITTI dataset, the experimental results verify the effectiveness of this method in small objects detection. Experimental results show that the proposed method significantly improves the detection capability of small objects compared with the baseline PointPillars. In concrete terms, in the moderate difficulty detection category, cars, pedestrians, and cyclists average precision (AP) values increased by 5.30%, 8.1%, and 10.6%, respectively. Moreover, the proposed method surpasses existing mainstream approaches in the cyclist category.

Oversized ore detection using rotated CCAD-YOLOv3 and ellipse fitting

The early detection and processing of oversized ores on the conveyor belt can effectively prevent accidents such as material blockage, ensuring transportation efficiency and production safety. The long-used manual detection method, however, has drawbacks such as high risk and poor accuracy. To address these challenges, this study proposes an automatic detection system for oversized ores based on rotated CCAD-YOLOv3 and ellipse fitting. Firstly, to more accurately represent the shape of adhesive ores and estimate their area, the rotated ellipse object detection is employed instead of conventional horizontal box object detection. Secondly, to enhance feature extraction capability and compensate for the low accuracy arising from insufficient data, the more sophisticated ConvNeXtv2-tiny with pre-trained parameters from ImageNet is adopted to replace DarkNet53 in the backbone. Then, to better integrate full-scale feature information, a structure combining the path aggregation network and adaptively spatial feature fusion with channel attention (PAN+CA-ASFF) is developed to substitute the feature pyramid network (FPN) in the neck. Lastly, to mitigate the inherent conflict between classification and regression tasks, a decoupling strategy is applied in the head. Experimental results demonstrate that the rotated ellipse is more conducive to fitting the shape of adhesive oversized ores than the segmentation mask, horizontal box, rotated box, and horizontal ellipse. Further, compared with the benchmark YOLOv3, CCAD-YOLOv3 exhibits a notable improvement in average precision (AP) by 6.52%, reaching 68.43%. This positions it as a leading solution, outperforming 21 other object detection algorithms.

Tiny drone object detection in videos guided by the bio-inspired magnocellular computation model

Detecting drones in infrared videos is highly desired in many realistic scenarios, e.g., unauthorized drone monitoring around airports. Nevertheless, automated drone detection is rather challenging when the targets appear as tiny objects (≤10×10 pixels) against complex backgrounds. Conventional object detection algorithms, which mainly use static visual features, can hardly distinguish tiny objects from undesired artefacts in complex backgrounds. To alleviate this problem, we learn from the early biological visual pathway (including the parvocellular and magnocellular pathways), which process static and motion information simultaneously. Therefore, we propose a magnocellular inspired method for video tiny-object detection (Magno-VTOD) that integrates both static and motion visual information. The Magno-VTOD firstly employs a retinal magnocellular computation model to extract the motion strength of moving objects. The motion responses are then used to enhance the areas of the flying tiny drones effectively and efficiently, thereby facilitating the subsequent target detection procedure. We implement the video tiny-object detection method based on the widely adopted deep neural networks guided by the magnocellular computation model. Experimental results obtained on the large-scale Anti-UAV dataset (304451 video frames) validate that the proposed Magno-VTOD method significantly outperforms the competing state-of-the-art object detection methods on the tiny drone detection task. Particularly, the AP value is increased by 15.4% for tiny object detection, and by 17.1%/13.7% against wood/mountain backgrounds.

CRTED: Few-Shot Object Detection via Correlation-RPN and Transformer Encoder–Decoder

Few-shot object detection (FSOD) aims to address the challenge of requiring a substantial number of annotations for training in conventional object detection, which is very labor-intensive. However, the existing few-shot methods achieve high precision with the sacrifice of time for exhaustive fine-tuning or have poor performance in novel-class adaptation. We presume the major reason is that the valuable correlation feature among different categories is insufficiently exploited, hindering the generalization of knowledge from base to novel categories for object detection. In this paper, we propose few-shot object detection via Correlation-RPN and transformer encoder–decoder (CRTED), a novel training network to learn object-relevant features of inter-class correlation and intra-class compactness while suppressing object-agnostic features in the background with limited annotated samples. And we also introduce a four-way tuple-contrast training strategy to positively activate the training progress of our object detector. Experiments over two few-shot benchmarks (Pascal VOC, MS-COCO) demonstrate that our proposed CRTED without further fine-tuning can achieve comparable performance with current state-of-the-art fine-tuned works. The codes and pre-trained models will be released.

Multiscale and multiperception feature learning for pancreatic lesion detection based on noncontrast CT

Background. Pancreatic cancer is one of the most malignant tumours, demonstrating a poor prognosis and nearly identically high mortality and morbidity, mainly because of the difficulty of early diagnosis and timely treatment for localized stages. Objective. To develop a noncontrast CT (NCCT)-based pancreatic lesion detection model that could serve as an intelligent tool for diagnosing pancreatic cancer early, overcoming the challenges associated with low contrast intensities and complex anatomical structures present in NCCT images. Approach. We design a multiscale and multiperception (MSMP) feature learning network with ResNet50 coupled with a feature pyramid network as the backbone for strengthening feature expressions. We added multiscale atrous convolutions to expand different receptive fields, contextual attention to perceive contextual information, and channel and spatial attention to focus on important channels and spatial regions, respectively. The MSMP network then acts as a feature extractor for proposing an NCCT-based pancreatic lesion detection model with image patches covering the pancreas as its input; Faster R-CNN is employed as the detection method for accurately detecting pancreatic lesions. Main results. By using the new MSMP network as a feature extractor, our model outperforms the conventional object detection algorithms in terms of the recall (75.40% and 90.95%), precision (40.84% and 68.21%), F1 score (52.98% and 77.96%), F2 score (64.48% and 85.26%) and Ap50 metrics (53.53% and 70.14%) at the image and patient levels, respectively. Significance.The good performance of our new model implies that MSMP can mine NCCT imaging features for detecting pancreatic lesions from complex backgrounds well. The proposed detection model is expected to be further developed as an intelligent method for the early detection of pancreatic cancer.

PL-DINO: An Improved Transformer-Based Method for Plant Leaf Disease Detection

Agriculture is important for ecology. The early detection and treatment of agricultural crop diseases are meaningful and challenging tasks in agriculture. Currently, the identification of plant diseases relies on manual detection, which has the disadvantages of long operation time and low efficiency, ultimately impacting the crop yield and quality. To overcome these disadvantages, we propose a new object detection method named “Plant Leaf Detection transformer with Improved deNoising anchOr boxes (PL-DINO)”. This method incorporates a Convolutional Block Attention Module (CBAM) into the ResNet50 backbone network. With the assistance of the CBAM block, the representative features can be effectively extracted from leaf images. Next, an EQualization Loss (EQL) is employed to address the problem of class imbalance in the relevant datasets. The proposed PL-DINO is evaluated using the publicly available PlantDoc dataset. Experimental results demonstrate the superiority of PL-DINO over the related advanced approaches. Specifically, PL-DINO achieves a mean average precision of 70.3%, surpassing conventional object detection algorithms such as Faster R-CNN and YOLOv7 for leaf disease detection in natural environments. In brief, PL-DINO offers a practical technology for smart agriculture and ecological monitoring.

Detection of Serrated Adenoma in NBI Based on Multi-Scale Sub-Pixel Convolution

Colorectal cancer ranks third in global malignancy incidence, and serrated adenoma is a precursor to colon cancer. However, current studies primarily focus on polyp detection, neglecting the crucial discrimination of polyp nature, hindering effective cancer prevention. This study established a static image dataset for serrated adenoma (SA) and developed a deep learning SA detection model. The proposed MSSDet (Multi-Scale Sub-pixel Detection) innovatively modifies each layer of the original feature pyramid’s structure to retain high-resolution polyp features. Additionally, feature fusion and optimization modules were incorporated to enhance multi-scale information utilization, leveraging the narrow-band imaging endoscope’s ability to provide clearer colonoscopy capillary and texture images. This paper utilized 639 cases of colonic NBI endoscopic images to construct the model, achieving a mean average precision (mAP) of 86.3% for SA in the test set. The SA detection rate via this approach has significantly surpassed conventional object detection methods.

Semi-supervised Open-World Object Detection

Conventional open-world object detection (OWOD) problem setting first distinguishes known and unknown classes and then later incrementally learns the unknown objects when introduced with labels in the subsequent tasks. However, the current OWOD formulation heavily relies on the external human oracle for knowledge input during the incremental learning stages. Such reliance on run-time makes this formulation less realistic in a real-world deployment. To address this, we introduce a more realistic formulation, named semi-supervised open-world detection (SS-OWOD), that reduces the annotation cost by casting the incremental learning stages of OWOD in a semi-supervised manner. We demonstrate that the performance of the state-of-the-art OWOD detector dramatically deteriorates in the proposed SS-OWOD setting. Therefore, we introduce a novel SS-OWOD detector, named SS-OWFormer, that utilizes a feature-alignment scheme to better align the object query representations between the original and augmented images to leverage the large unlabeled and few labeled data. We further introduce a pseudo-labeling scheme for unknown detection that exploits the inherent capability of decoder object queries to capture object-specific information. On the COCO dataset, our SS-OWFormer using only 50% of the labeled data achieves detection performance that is on par with the state-of-the-art (SOTA) OWOD detector using all the 100% of labeled data. Further, our SS-OWFormer achieves an absolute gain of 4.8% in unknown recall over the SOTA OWOD detector. Lastly, we demonstrate the effectiveness of our SS-OWOD problem setting and approach for remote sensing object detection, proposing carefully curated splits and baseline performance evaluations. Our experiments on 4 datasets including MS COCO, PASCAL, Objects365 and DOTA demonstrate the effectiveness of our approach. Our source code, models and splits are available here https://github.com/sahalshajim/SS-OWFormer

DoubleAUG: Single-domain Generalized Object Detector in Urban via Color Perturbation and Dual-style Memory

Object detection in urban scenarios is crucial for autonomous driving in intelligent traffic systems. However, unlike conventional object detection tasks, urban-scene images vary greatly in style. For example, images taken on sunny days differ significantly from those taken on rainy days. Therefore, models trained on sunny-day images may not generalize well to rainy-day images. In this article, we aim to solve the single-domain generalizable object detection task in urban scenarios, meaning that a model trained on images from one weather condition should be able to perform well on images from any other weather conditions. To address this challenge, we propose a novel Double AUGmentation (DoubleAUG) method that includes image- and feature-level augmentation schemes. In the image-level augmentation, we consider the variation in color information across different weather conditions and propose a Color Perturbation (CP) method that randomly exchanges the RGB channels to generate various images. In the feature-level augmentation, we propose to utilize a Dual-Style Memory (DSM) to explore the diverse style information on the entire dataset, further enhancing the model’s generalization capability. Extensive experiments demonstrate that our proposed method outperforms state-of-the-art methods. Furthermore, ablation studies confirm the effectiveness of each module in our proposed method. Moreover, our method is plug-and-play and can be integrated into existing methods to further improve model performance.

CETR: CenterNet-Vision transformer model for wheat head detection

<p>Wheat head detection is a critical task in precision agriculture for estimating crop yield and optimizing agricultural practices. Conventional object detection architectures often struggle with detecting densely packed and overlapping wheat heads in complex agricultural field images. To address this challenge, a novel CEnternet-vision TRansformer model for Wheat Head Detection (CETR) is proposed. CETR model combines the strengths of two cutting-edge technologies—CenterNet and Vision Transformer. A dataset of agricultural farm images labeled with precise wheat head annotations is used to train and evaluate the CETR model. Comprehensive experiments were conducted to compare CETR’s performance against convolutional neural network model commonly used in agricultural applications. The higher mAP value of 0.8318 for CETR compared against AlexNet, VGG19, ResNet152 and MobileNet indicates that the CETR model is more effective in detecting wheat heads in agricultural images. It achieves a higher precision in predicting bounding boxes that align well with the ground truth, resulting in more accurate and reliable wheat head detection. The higher performance of CETR can be attributed to the combination of CenterNet and ViT as a two-stage architecture taking advantage of both methods. Moreover, the transformer-based architecture of CETR enables better generalization across different agricultural environments, making it a suitable solution for automated agricultural applications.</p>

Enabling Visual Object Detection With Object Sounds via Visual Modality Recalling Memory.

When humans hear the sound of an object, they recall associated visual information and integrate the sound with recalled visual modality to detect the object. In this article, we present a novel sound-based object detector that mimics this process. We design a visual modality recalling (VMR) memory to recall information of a visual modality based on an audio modal input (i.e., sound). To achieve this goal, we propose a VMR loss and an audio-visual association loss to guide the VMR memory to memorize visual modal information by establishing associations between audio and visual modalities. With the visual modal information recalled through the VMR memory along with the original audio input, we perform audio-visual integration. In this step, we introduce an integrated feature contrastive loss that allows the integrated feature to be embedded as if it were encoded using both audio and visual modal inputs. This guidance enables our sound-based object detector to effectively perform visual object detection even when only sound is provided. We believe that our work is a cornerstone study that offers a new perspective to conventional object detection studies that solely rely on the visual modality. Comprehensive experimental results demonstrate the effectiveness of the proposed method with the VMR memory.

An Innovative Deep Learning Approach to Spinal Fracture Detection in CT Images.

Spinal fractures, particularly vertebral compression fractures, pose a significant challenge in medical imaging due to their small-scale nature and blurred boundaries in Computed Tomography (CT) scans. However, advanced deep learning models, such as the integration of the You Only Look Once (YOLO) V7 model with Efficient Layer Aggregation Networks (ELAN) and Max-Pooling Convolution (MPConv) architectures, can substantially reduce the loss of small-scale information during computational processing, thus improving detection accuracy. The purpose of this study is to develop an innovative deep learning approach for detecting spinal fractures, particularly vertebral compression fractures, in CT images. We proposed a novel method to precisely identify spinal injury using the YOLO V7 model as a classifier. This model was enhanced by integrating ELAN and MPConv architectures, which were influenced by the Receptive Field Learning and Aggregation (RFLA) small object recognition framework. Standard normalization techniques were utilized to preprocess the CT images. The YOLO V7 model, integrated with ELAN and MPConv architectures, was trained using a dataset containing annotated spinal fractures. Additionally, to mitigate boundary ambiguities in compressive fractures, a Theoretical Receptive Field (TRF) based on Gaussian distribution and an Effective Receptive Field (ERF) were used to capture multi-scale features better. Furthermore, the Wasserstein distance was employed to optimize the model's learning process. A total of 240 CT images from patients diagnosed with spinal fractures were included in this study, sourced from Ningbo No.2 Hospital, ensuring a robust dataset for training the deep learning model. Our method demonstrated superior performance over conventional object detection networks like YOLO V7 and YOLO V3. Specifically, with a dataset of 200 pathological images and 40 normal spinal images, our method achieved a 3% increase in accuracy compared to YOLO V7. The proposed method offers an innovative and more effective approach for identifying vertebral compression fractures in CT scans. These promising findings suggest the method's potential for practical clinical applications, highlighting the significance of deep learning in enhancing patient care and treatment in medical imaging. Future research should incorporate cross-validation and independent validation and test sets to assess the model's robustness and generalizability. Additionally, exploring other deep learning models and methods could further enhance detection accuracy and reliability, contributing to the development of more effective diagnostic tools in medical imaging.

Polarity Loss: Improving Visual-Semantic Alignment for Zero-Shot Detection.

Conventional object detection models require large amounts of training data. In comparison, humans can recognize previously unseen objects by merely knowing their semantic description. To mimic similar behavior, zero-shot object detection (ZSD) aims to recognize and localize "unseen" object instances by using only their semantic information. The model is first trained to learn the relationships between visual and semantic domains for seen objects, later transferring the acquired knowledge to totally unseen objects. This setting gives rise to the need for correct alignment between visual and semantic concepts so that the unseen objects can be identified using only their semantic attributes. In this article, we propose a novel loss function called "polarity loss" that promotes correct visual-semantic alignment for an improved ZSD. On the one hand, it refines the noisy semantic embeddings via metric learning on a "semantic vocabulary" of related concepts to establish a better synergy between visual and semantic domains. On the other hand, it explicitly maximizes the gap between positive and negative predictions to achieve better discrimination between seen, unseen, and background objects. Our approach is inspired by embodiment theories in cognitive science that claim human semantic understanding to be grounded in past experiences (seen objects), related linguistic concepts (word vocabulary), and visual perception (seen/unseen object images). We conduct extensive evaluations on the Microsoft Common Objects in Context (MS-COCO) and Pascal Visual Object Classes (VOC) datasets, showing significant improvements over state of the art. Our code and evaluation protocols available at: https://github.com/salman-h-khan/PL-ZSD_Release.

Backlight and dim space object detection based on a novel event camera.

For space object detection tasks, conventional optical cameras face various application challenges, including backlight issues and dim light conditions. As a novel optical camera, the event camera has the advantages of high temporal resolution and high dynamic range due to asynchronous output characteristics, which provides a new solution to the above challenges. However, the asynchronous output characteristic of event cameras makes them incompatible with conventional object detection methods designed for frame images. Asynchronous convolutional memory network (ACMNet) for processing event camera data is proposed to solve the problem of backlight and dim space object detection. The key idea of ACMNet is to first characterize the asynchronous event streams with the Event Spike Tensor (EST) voxel grid through the exponential kernel function, then extract spatial features using a feed-forward feature extraction network, and aggregate temporal features using a proposed convolutional spatiotemporal memory module ConvLSTM, and finally, the end-to-end object detection using continuous event streams is realized. Comparison experiments among ACMNet and classical object detection methods are carried out on Event_DVS_space7, which is a large-scale space synthetic event dataset based on event cameras. The results show that the performance of ACMNet is superior to the others, and the mAP is improved by 12.7% while maintaining the processing speed. Moreover, event cameras still have a good performance in backlight and dim light conditions where conventional optical cameras fail. This research offers a novel possibility for detection under intricate lighting and motion conditions, emphasizing the superior benefits of event cameras in the realm of space object detection.

Small Target-YOLOv5: Enhancing the Algorithm for Small Object Detection in Drone Aerial Imagery Based on YOLOv5.

Object detection in drone aerial imagery has been a consistent focal point of research. Aerial images present more intricate backgrounds, greater variation in object scale, and a higher occurrence of small objects compared to standard images. Consequently, conventional object detection algorithms are often unsuitable for direct application in drone scenarios. To address these challenges, this study proposes a drone object detection algorithm model based on YOLOv5, named SMT-YOLOv5 (Small Target-YOLOv5). The enhancement strategy involves improving the feature fusion network by incorporating detection layers and implementing a weighted bidirectional feature pyramid network. Additionally, the introduction of the Combine Attention and Receptive Fields Block (CARFB) receptive field feature extraction module and DyHead dynamic target detection head aims to broaden the receptive field, mitigate information loss, and enhance perceptual capabilities in spatial, scale, and task domains. Experimental validation on the VisDrone2021 dataset confirms a significant improvement in the target detection accuracy of SMT-YOLOv5. Each improvement strategy yields effective results, raising the average precision by 12.4 percentage points compared to the original method. Detection improvements for large, medium, and small targets increase by 6.9%, 9.5%, and 7.7%, respectively, compared to the original method. Similarly, applying the same improvement strategies to the low-complexity YOLOv8n results in SMT-YOLOv8n, which is comparable in complexity to SMT-YOLOv5s. The results indicate that, relative to SMT-YOLOv8n, SMT-YOLOv5s achieves a 2.5 percentage point increase in average precision. Furthermore, comparative experiments with other enhancement methods demonstrate the effectiveness of the improvement strategies.

BA-YOLO for Object Detection in Satellite Remote Sensing Images

In recent years, there has been significant progress in object detection within the domain of natural images. However, the field of satellite remote sensing images has consistently presented challenges due to its significant scale variations and complex background interference. Achieving satisfactory results by directly applying conventional image object detection models has proven to be difficult. To address these challenges, this paper introduces BA-YOLO, an improved version of the YOLOv8 object detection model. It incorporates several notable enhancements. Firstly, to fuse an increased number of features more effectively, we introduce the design concept of a higher-performing Bi-directional Feature Pyramid Network (BiFPN). Secondly, to retain sufficient global contextual information, we integrated a module in BA-YOLO that combines multi-head self-attention and convolutional networks. Finally, we employed various data augmentation techniques such as Mixup, Cutout, Mosaic, and multi-scale training to enhance the model’s accuracy and robustness. Experimental results demonstrate that BA-YOLO outperforms state-of-the-art detectors and has been evaluated on the DOTA dataset. BA-YOLO achieves a mean average precision (mAP) of 0.722 on the DOTA dataset.

Satellite–Terrestrial Collaborative Object Detection via Task-Inspired Framework

Recently, buoyed by advances in the space industry, low Earth orbit (LEO) satellites have become an important part of the Internet of Things (IoT). LEO satellites have entered the era of a big data link with IoT, how to deal with the data from the satellite IoT is a problem worthy of consideration. Conventional object detection method in optical remote sensing simply transmits the raw data to the ground. However, it ignores the properties of the images and the connection with the downstream task. To obtain efficient data transmission and accurate object detection, we propose a task-inspired satellite-terrestrial collaborative object detection framework called STCOD. It detects regions of interest (ROIs) and adopts a block-based adaptive sampling method to compress the background (BG) in optical remote sensing images by introducing satellite edge computing (SEC) on satellites. The STCOD framework also sets the transmission priority of image blocks according to their contributions to the task and uses fountain code to ensure the reliable transmission of important image blocks. We build a whole software simulation framework to validate our method, including the satellite module, the transmission module, and the terrestrial module. Extensive experimental results show that the STCOD framework can reduce the amount of downlink data decreased by 50.04% while losing the detection accuracy by 0.54%. In our simulated satellite-terrestrial link, the STCOD framework can reduce the number of satellite-to-terrestrial transmissions by half. When the packet loss rate is between 5% and 20%, the detection accuracy is lost only 0.05% to 0.5%.