MegaDetectNet: A Fast Object Detection Framework for Ultra-High-Resolution Images

Deep convolutional neural network (DCNN) based computer vision methods have great progress in object detection tasks. And object detection is essential to many applications such as autonomous robots and vehicles. However, it is difficult for object detection system to be deployed on embedded platforms due to its intensive computing architecture. We present an object detector MobileNet-SSD which can be deployed on embedded platforms. The object detector presented in this work based on Single Shot Detector (SSD) framework and replace the feature extractor with a more light weight network MobileNetV1. General matrix to matrix multiplication is used to simplify the calculation of convolution operation and is further optimized using ARM NEON technology. Finally, the object detector MobileNet-SSD is deployed on embedded platform and achieves 1.13FPS with 0.72 mean average precision on PASCAL VOC dataset.

Object detection and recognition is an essential task in computer vision with numerous real-world applications such as surveillance, self-driving cars, and robotics. In recent years, deep learning-based approaches have significantly improved the accuracy and speed of object detection and recognition. The You Only Look Once version 3 (YOLOv3) algorithm is a popular deep learning-based approach that can detect and recognize objects in real-time. The Common Objects in Context (COCO) dataset is a large-scale dataset with over 330,000 labeled images and more than 2.5 million object instances, making it a popular choice for object detection and recognition tasks. In this paper, we propose a deep learning-based approach for real-time object detection and recognition using the YOLOv3 architecture and COCO dataset. We evaluate our approach based on several performance metrics, including mean average precision (mAP), frames per second (FPS), total object detection time, object detection accuracy, false positive rate, number of detected objects, and mean intersection over union (mIoU). Our results show that our approach achieves a mean average precision of 0.76 on the COCO dataset and a real-time performance of 40 frames per second on a single GPU. Additionally, our approach achieves an object detection accuracy of 93.5%, a false positive rate of 6.5%, and a mean intersection over union of 0.65. Our proposed approach shows promising results for real-time object detection and recognition and can be applied to various real-world applications.

Generally, most approaches using methods such as cropping, rotating, and flipping achieve more data to train models for improving the accuracy of detection and segmentation. However, due to the difficulties of labeling such data especially semantic segmentation data, those traditional data augmentation methodologies cannot help a lot when the training set is really limited. In this paper, a model named OFA-Net (One For All Network) is proposed to combine object detection and semantic segmentation tasks. Meanwhile, using a strategy called “1-N Alternation” to train the OFA-Net model, which can make a fusion of features from detection and segmentation data. The results show that object detection data can be recruited to better the segmentation accuracy performance, and furthermore, segmentation data assist a lot to enhance the confidence of predictions for object detection. Finally, the OFA-Net model is trained without traditional data augmentation methodologies and tested on the KITTI test server. The model works well on the KITTI Road Segmentation challenge and can do a good job on the object detection task.

Deep neural network approaches have demonstrated high performance in object recognition (CNN) and detection (Faster-RCNN) tasks, but experiments have shown that such architectures are vulnerable to adversarial attacks (FFF, UAP): low amplitude perturbations, barely perceptible by the human eye, can lead to a drastic reduction in labeling performance. This article proposes a new context module, called \textit{Transformer-Encoder Detector Module}, that can be applied to an object detector to (i) improve the labeling of object instances; and (ii) improve the detector's robustness to adversarial attacks. The proposed model achieves higher mAP, F1 scores and AUC average score of up to 13\% compared to the baseline Faster-RCNN detector, and an mAP score 8 points higher on images subjected to FFF or UAP attacks due to the inclusion of both contextual and visual features extracted from scene and encoded into the model. The result demonstrates that a simple ad-hoc context module can improve the reliability of object detectors significantly.

Due to high-speed motion blur and challenging illumination, conventional frame-based cameras have encountered an important challenge in object detection tasks. Neuromorphic cameras that output asynchronous visual streams instead of intensity frames, by taking the advantage of high temporal resolution and high dynamic range, have brought a new perspective to address the challenge. In this paper, we propose a novel problem setting, retinomorphic object detection, which is the first trial that integrates foveal-like and peripheral-like visual streams. Technically, we first build a large-scale multimodal neuromorphic object detection dataset (i.e., PKU-Vidar-DVS) over 215.5k spatio-temporal synchronized labels. Then, we design temporal aggregation representations to preserve the spatio-temporal information from asynchronous visual streams. Finally, we present a novel bio-inspired unifying framework to fuse two sensing modalities via a dynamic interaction mechanism. Our experimental evaluation shows that our approach has significant improvements over the state-of-the-art methods with the single-modality, especially in high-speed motion and low-light scenarios. We hope that our work will attract further research into this newly identified, yet crucial research direction. Our dataset can be available at https://www.pkuml.org/resources/pku-vidar-dvs.html.

Object localization and detection is the emerging area of computer vision and image processing that finds out the location of the object in the video frames or in the digital images. Object localization and detection have many challenges like occlusion, scale variations, intraclass similarities, illumination conditions, pose variations, etc. Multi-view object detection focuses on different views of objects like the top, bottom, side view, etc. We aim to discuss the different Object detection models based on Machine learning and Deep neural network approaches in a multi-view environment. We also discuss different datasets used for object detections and their applications. We apply some object detection models like SSD and Faster R-CNN to specific object categories of Open Image Dataset V6 and compare the results based on mAP (Mean Average Precision).KeywordsObject detectionConvolution neural networksR-CNNFast R-CNNFaster R-CNNEfficientDetCenterNet

Object detection is an essential and crucial task in remote sensing images. However, its performance is often constrained due to the multiple rotation angles and small size of the objects. To solve this issue, this letter proposes an object detection algorithm called ODC-YOLO, which is built based on the YOLOv5. Firstly, we integrate Omni-Dimensional dynamic convolution (ODConv) with Res2Net to reconstruct the backbone, generating four types of attention mechanisms to extract feature information from four dimensions of convolution layers, enhancing the detection capability for small objects; secondly, we propose a Multi-Scale Receptive Filed Block (M-RFB) for discrimination between similar objects; finally, this letter use the Circular Smooth Labels (CSL) converts angle regression into a classification problem, and replace horizontal bounding boxes with oriented bounding boxes, addressing the issues of diverse object angles and dense arrangement. Verified through experiments, ODC-YOLO demonstrates an increase of 3.45% in mean Average Precision (mAP) on DOTA dataset, compared with YOLOv10n, meanwhile, the mAP for small object reached 26.5%.

Object detection from waterborne imagery is an essential aspect in maritime traffic management, navigation safety and coastal security. Building efficient autonomous systems, which can take decisions in critical situations, requires an integration of complex object detectors in real time. Object detectors trained on generic datasets often give unsatisfactory results in complex scenarios like the maritime environment, since only a fraction of their images contain maritime vessels. Publicly available domain-specific datasets are scarce, and they are limited in the number of images and annotations. Compared to object detection in applications such as autonomous vehicles, maritime vessel detection is considerably reduced in computer vision research. This creates a deficit in exhaustive benchmarking studies for maritime detection datasets. To bridge this gap, we relabel the ABOships dataset and benchmark a state-of-the-art center-based detector, Centernet, on the newly relabeled dataset, ABOships-PLUS. We explore its performance under different feature extractors, and investigate the effect of object size and inter-class variation on detection accuracy. The reported benchmarking illustrates that the ABOships-PLUS dataset is adequate to use in supervised domain adaptation. The experimental results show that Centernet with DLA (Deep Layer Aggregation) as a feature extractor achieved the highest accuracy in detecting maritime objects overall (with mean average precision 74.4%).

In recent years, Convolutional Neural Networks (CNNs) have repeatedly shown state-of-the-art performance for their accuracy in the task of object detection, but their heavy computational costs impede their ability for real-time detection when the supporting system is moving, particulary when it is accelerating. At the same time, recent progress on visual inertial systems takes great advantage of movement information to robustly estimate the robot state and its surrounding. This paper proposes to exploit the advantages of inertial odometry research for the purpose of real-time object detection system on mobile robots. We combine a CNN detector with VINS-Mono, a moving visual odometry system, and show reliable improvement in the detection process, especially when the robot accelerates or decelerates. Our system is ready-to-use in that it has very low deployment cost and requires no calibration. The resulting system allows for simultaneous robot state estimation and object detection, as well as object tracking. Lastly, this architecture proves to be flexible because not restrained to a specific object type or detector.

<p>The ultimate goal of this study is to afford enhanced video object detection and tracking by eliminating the limitations which are existing nowadays. Although high performance ratio for video object detection and tracking is achieved in the earlier work it takes more time for computation. Consequently we are in need to propose a novel video object detection and tracking technique so as to minimize the computational complexity. Our proposed technique covers five stages they are preprocessing, segmentation, feature extraction, background subtraction and hole filling. Originally the video clip in the database is split into frames. Then preprocessing is performed so as to get rid of noise, an adaptive median filter is used in this stage to eliminate the noise. The preprocessed image then undergoes segmentation by means of modified region growing algorithm. The segmented image is subjected to feature extraction phase so as to extract the multi features from the segmented image and the background image, the feature value thus obtained are compared so as to attain optimal value, consequently a foreground image is attained in this stage. The foreground image is then subjected to morphological operations of erosion and dilation so as to fill the holes and to get the object accurately as these foreground image contains holes and discontinuities. Thus the moving object is tracked in this stage. This method will be employed in MATLAB platform and the outcomes will be studied and compared with the existing techniques so as to reveal the performance of the novel video object detection and tracking technique. </p>

Object detection is a fundamental problem in computer vision, with applications spanning self-driving cars, surveillance systems, medical imaging, robotics, and smart cities. Among the myriad of algorithms developed for this task, the You Only Look Once (YOLO) family stands out for its ability to perform real-time and accurate object detection. This article provides a comprehensive analysis of the YOLO algorithm series, from YOLOv1 to YOLOv8, evaluating them across key performance metrics, including precision, recall, mean Average Precision (mAP), frames per second (FPS), and overall effectiveness. Unlike traditional two-stage detectors such as R-CNN, YOLO formulates object detection as a single regression problem: a single pass over the image simultaneously predicts bounding boxes and class probabilities. This end-to-end design enables YOLO to achieve high speed while maintaining a competitive accuracy-efficiency trade-off. We examine architectural innovations across YOLO versions, including batch normalization, anchor boxes, residual blocks, feature pyramid networks, and attention mechanisms, and discuss their impact on performance. Lightweight models (e.g., YOLOv5-Nano, YOLOv8-Small) are explored with a focus on their suitability for mobile and embedded systems, highlighting YOLO’s adaptability to resource-constrained environments. Challenges such as small object detection, occlusion, and domain-specific tuning are also addressed. This article serves as a practical guide for researchers, developers, and practitioners aiming to leverage YOLO for real-world object detection tasks.

As the core technologies represented by Internet of Things (IoT) technology and computer vision technology booming, object detection algorithms based on deep learning have received high attention in various fields. Object detection is not only a core problem in the field of computer vision, but also a prerequisite and basis for many computer vision tasks, and has important applications in the fields of autonomous driving and video surveillance. The task of object detection is to find all the targets of interest in an image and determine their category and location. Object detection includes two tasks: object classification and object localization. However, object classification focuses on the most discriminative part of the feature map, and object localization requires a feature map focusing on the whole region of the object. Therefore, in order to improve the detection accuracy, we propose a CBAM-RetinaNet network model.CBAM (Convolutional Block Attention Module) is a lightweight convolutional attention module, which combines channel and spatial attention mechanism modules. The channel attention module focuses on the meaningful information in the input image, which would lead to an increase in the accuracy of object classification. The spatial attention module focuses on the location information of the object, which may improve the accuracy of object localization. The experiment results show that the proposed achieves better object detection performance on the PASCAL VOC2007 dataset, and the accuracy is 2% higher than that of the RetinaNet model, which has good experimental results.

The problem of image noise is prevalent in various visual recognition tasks. This problem arises due to multiple factors like insufficient illumination, low-quality camera sensors, background-clutters, etc. Low-quality images incur degraded performance for visual recognition tasks like classification, object detection, because of the noise present in the image. Thus, mostly the latest image recognition benchmarks and methods, such as Pascal Visual Object Classes Challenges and Microsoft Common Objects in Context Challenges, rely primarily on clean and high-definition images. Furthermore, image noise is a typical problem in real-time applications like surveillance, autonomous vehicles, etc. To tackle this bothersome problem of underperformance of visual recognition tasks, particularly object detection, due to inclusion of noise in real-time scenarios, a technique has been proposed to reduce the image noise's impact on the object detection task and improve its performance altogether. A two-phase pipeline is introduced, in which the first phase consists ofremoving noise from the low-quality image, and the second stage includes accomplishing the object detection process by employing the standard Single-shot multibox object detector. The proposed approach is evaluated on the Pascal Visual Object Classes benchmark. The detection results and image denoising results exhibit the efficiency of the proposed technique for reliable object detection with varying image noise levels.

Drones, or general UAVs, equipped with cameras have been fast deployed with a wide range of applications, including agriculture, aerial photography, and surveillance. Consequently, automatic understanding of visual data collected from drones becomes highly demanding, bringing computer vision and drones more and more closely. To promote and track the developments of object detection and tracking algorithms, we have organized three challenge workshops in conjunction with ECCV 2018, ICCV 2019 and ECCV 2020, attracting more than 100 teams around the world. We provide a large-scale drone captured dataset, VisDrone, which includes four tracks, i.e., (1) image object detection, (2) video object detection, (3) single object tracking, and (4) multi-object tracking. In this paper, we first present a thorough review of object detection and tracking datasets and benchmarks, and discuss the challenges of collecting large-scale drone-based object detection and tracking datasets with fully manual annotations. After that, we describe our VisDrone dataset, which is captured over various urban/suburban areas of 14 different cities across China from North to South. Being the largest such dataset ever published, VisDrone enables extensive evaluation and investigation of visual analysis algorithms for the drone platform. We provide a detailed analysis of the current state of the field of large-scale object detection and tracking on drones, and conclude the challenge as well as propose future directions. We expect the benchmark largely boost the research and development in video analysis on drone platforms. All the datasets and experimental results can be downloaded from https://github.com/VisDrone/VisDrone-Dataset.

Object tracking can be considered as an object detection task. However, in general, an object detector cannot differentiate among intraclass instances and cannot fully exploit the spatiotemporal clues between successive frames. In order to solve these problems, in this paper, we proposed a novel object tracking framework based on object detection. We name the new framework TSTrack. Firstly, transformer is used in the target guidance module to extract the object’s features to guide the backbone network to complete the positioning task, which can solve the problem of distinguishing the instances within the classification. Secondly, self-attention mechanism is introduced into the model, which is mainly completed by spatial attention module (SAM) and channel attention module (CAM). By using the characteristics of self-attention mechanism, the attention matrix is dynamically self-learned in two dimensions of space and channel to obtain a more robust object representation. The proposed model is evaluated on the OTB-2013, OTB-50, OTB-2015 and VOT-2016 benchmarks, and the experimental results demonstrate the excellent tracking performance and robustness of our model.