Object detectors have demonstrated vulnerability to adversarial examples crafted by small perturbations that can deceive the object detector. Existing adversarial attacks mainly focus on white-box attacks and are merely valid at a specific viewpoint, while the universal multi-view black-box attack is less explored, limiting their generalization in practice. In this paper, we propose a novel universal multi-view black-box attack against object detectors, which optimizes a universal adversarial UV texture constructed by multiple image stickers for a 3D object via the designed layout optimization algorithm. Specifically, we treat the placement of image stickers on the UV texture as a circle-based layout optimization problem, whose objective is to find the optimal circle layout filled with image stickers so that it can deceive the object detector under the multi-view scenario. To ensure reasonable placement of image stickers, two constraints are elaborately devised. To optimize the layout, we adopt the random search algorithm enhanced by the devised important-aware selection strategy to find the most appropriate image sticker for each circle from the image sticker pools. Extensive experiments conducted on four common object detectors suggested that the detection performance decreases by a large magnitude of 74.29% on average in multi-view scenarios. Additionally, a novel evaluation tool based on the photo-realistic simulator is designed to assess the texture-based attack fairly.
Intracerebral hemorrhage is one of the diseases with the highest mortality and poorest prognosis worldwide. Spontaneous intracerebral hemorrhage (SICH) typically presents acutely, prompt and expedited radiological examination is crucial for diagnosis, localization, and quantification of the hemorrhage. Early detection and accurate segmentation of perihematomal edema (PHE) play a critical role in guiding appropriate clinical intervention and enhancing patient prognosis. However, the progress and assessment of computer-aided diagnostic methods for PHE segmentation and detection face challenges due to the scarcity of publicly accessible brain CT image datasets. This study establishes a publicly available CT dataset named PHE-SICH-CT-IDS for perihematomal edema in spontaneous intracerebral hemorrhage. The dataset comprises 120 brain CT scans and 7,022 CT images, along with corresponding medical information of the patients. To demonstrate its effectiveness, classical algorithms for semantic segmentation, object detection, and radiomic feature extraction are evaluated. The experimental results confirm the suitability of PHE-SICH-CT-IDS for assessing the performance of segmentation, detection and radiomic feature extraction methods. To the best of our knowledge, this is the first publicly available dataset for PHE in SICH, comprising various data formats suitable for applications across diverse medical scenarios. We believe that PHE-SICH-CT-IDS will allure researchers to explore novel algorithms, providing valuable support for clinicians and patients in the clinical setting. PHE-SICH-CT-IDS is freely published for non-commercial purpose at: https://figshare.com/articles/dataset/PHE-SICH-CT-IDS/23957937.
In recent years, object detection in deep learning has experienced rapid development. However, most existing object detection models perform well only on closed-set datasets, ignoring a large number of potential objects whose categories are not defined in the training set. These objects are often identified as background or incorrectly classified as pre-defined categories by the detectors. In this paper, we focus on the challenging problem of Novel Class Discovery and Localization (NCDL), aiming to train detectors that can detect the categories present in the training data, while also actively discover, localize, and cluster new categories. We analyze existing NCDL methods and identify the core issue: object detectors tend to be biased towards seen objects, and this leads to the neglect of unseen targets. To address this issue, we first propose an Debiased Region Mining (DRM) approach that combines class-agnostic Region Proposal Network (RPN) and class-aware RPN in a complementary manner. Additionally, we suggest to improve the representation network through semi-supervised contrastive learning by leveraging unlabeled data. Finally, we adopt a simple and efficient mini-batch K-means clustering method for novel class discovery. We conduct extensive experiments on the NCDL benchmark, and the results demonstrate that the proposed DRM approach significantly outperforms previous methods, establishing a new state-of-the-art.
The detection of polylines is usually either bound to branchless polylines or formulated in a recurrent way, prohibiting their use in real-time systems. We propose an approach that builds upon the idea of single shot object detection. Reformulating the problem of polyline detection as a bottom-up composition of small line segments allows to detect bounded, dashed and continuous polylines with a single head. This has several major advantages over previous methods. Not only is the method at 187 fps more than suited for real-time applications with virtually any restriction on the shapes of the detected polylines. By predicting multiple line segments for each cell, even branching or crossing polylines can be detected. We evaluate our approach on three different applications for road marking, lane border and center line detection. Hereby, we demonstrate the ability to generalize to different domains as well as both implicit and explicit polyline detection tasks.
Road obstacle detection is an important problem for vehicle driving safety. In this paper, we aim to obtain robust road obstacle detection based on spatio-temporal context modeling. Firstly, a data-driven spatial context model of the driving scene is constructed with the layouts of the training data. Then, obstacles in the input image are detected via the state-of-the-art object detection algorithms, and the results are combined with the generated scene layout. In addition, to further improve the performance and robustness, temporal information in the image sequence is taken into consideration, and the optical flow is obtained in the vicinity of the detected objects to track the obstacles across neighboring frames. Qualitative and quantitative experiments were conducted on the Small Obstacle Detection (SOD) dataset and the Lost and Found dataset. The results indicate that our method with spatio-temporal context modeling is superior to existing methods for road obstacle detection.
The classification and regression head are both indispensable components to build up a dense object detector, which are usually supervised by the same training samples and thus expected to have consistency with each other for detecting objects accurately in the detection pipeline. In this paper, we break the convention of the same training samples for these two heads in dense detectors and explore a novel supervisory paradigm, termed as Mutual Supervision (MuSu), to respectively and mutually assign training samples for the classification and regression head to ensure this consistency. MuSu defines training samples for the regression head mainly based on classification predicting scores and in turn, defines samples for the classification head based on localization scores from the regression head. Experimental results show that the convergence of detectors trained by this mutual supervision is guaranteed and the effectiveness of the proposed method is verified on the challenging MS COCO benchmark. We also find that tiling more anchors at the same location benefits detectors and leads to further improvements under this training scheme. We hope this work can inspire further researches on the interaction of the classification and regression task in detection and the supervision paradigm for detectors, especially separately for these two heads.
We present YOLOBench, a benchmark comprised of 550+ YOLO-based object detection models on 4 different datasets and 4 different embedded hardware platforms (x86 CPU, ARM CPU, Nvidia GPU, NPU). We collect accuracy and latency numbers for a variety of YOLO-based one-stage detectors at different model scales by performing a fair, controlled comparison of these detectors with a fixed training environment (code and training hyperparameters). Pareto-optimality analysis of the collected data reveals that, if modern detection heads and training techniques are incorporated into the learning process, multiple architectures of the YOLO series achieve a good accuracy-latency trade-off, including older models like YOLOv3 and YOLOv4. We also evaluate training-free accuracy estimators used in neural architecture search on YOLOBench and demonstrate that, while most state-of-the-art zero-cost accuracy estimators are outperformed by a simple baseline like MAC count, some of them can be effectively used to predict Pareto-optimal detection models. We showcase that by using a zero-cost proxy to identify a YOLO architecture competitive against a state-of-the-art YOLOv8 model on a Raspberry Pi 4 CPU. The code and data are available at https://github.com/Deeplite/deeplite-torch-zoo
Explainable artificial intelligence (XAI) has witnessed significant advances in the field of object recognition, with saliency maps being used to highlight image features relevant to the predictions of learned models. Although these advances have made AI-based technology more interpretable to humans, several issues have come to light. Some approaches present explanations irrelevant to predictions, and cannot guarantee the validity of XAI (axioms). In this study, we propose the Baseline Shapley-based Explainable Detector (BSED), which extends the Shapley value to object detection, thereby enhancing the validity of interpretation. The Shapley value can attribute the prediction of a learned model to a baseline feature while satisfying the explainability axioms. The processing cost for the BSED is within the reasonable range, while the original Shapley value is prohibitively computationally expensive. Furthermore, BSED is a generalizable method that can be applied to various detectors in a model-agnostic manner, and interpret various detection targets without fine-grained parameter tuning. These strengths can enable the practical applicability of XAI. We present quantitative and qualitative comparisons with existing methods to demonstrate the superior performance of our method in terms of explanation validity. Moreover, we present some applications, such as correcting detection based on explanations from our method.
Few-shot object detection~(FSOD), which aims to detect novel objects with limited annotated instances, has made significant progress in recent years. However, existing methods still suffer from biased representations, especially for novel classes in extremely low-shot scenarios. During fine-tuning, a novel class may exploit knowledge from similar base classes to construct its own feature distribution, leading to classification confusion and performance degradation. To address these challenges, we propose a fine-tuning based FSOD framework that utilizes semantic embeddings for better detection. In our proposed method, we align the visual features with class name embeddings and replace the linear classifier with our semantic similarity classifier. Our method trains each region proposal to converge to the corresponding class embedding. Furthermore, we introduce a multimodal feature fusion to augment the vision-language communication, enabling a novel class to draw support explicitly from well-trained similar base classes. To prevent class confusion, we propose a semantic-aware max-margin loss, which adaptively applies a margin beyond similar classes. As a result, our method allows each novel class to construct a compact feature space without being confused with similar base classes. Extensive experiments on Pascal VOC and MS COCO demonstrate the superiority of our method.
Identifying and localizing objects within images is a fundamental challenge, and numerous efforts have been made to enhance model accuracy by experimenting with diverse architectures and refining training strategies. Nevertheless, a prevalent limitation in existing models is overemphasizing the current input while ignoring the information from the entire dataset. We introduce an innovative {\em \textbf{R}etriever}-{\em\textbf{D}ictionary} (RD) module to address this issue. This architecture enables YOLO-based models to efficiently retrieve features from a Dictionary that contains the insight of the dataset, which is built by the knowledge from Visual Models (VM), Large Language Models (LLM), or Visual Language Models (VLM). The flexible RD enables the model to incorporate such explicit knowledge that enhances the ability to benefit multiple tasks, specifically, segmentation, detection, and classification, from pixel to image level. The experiments show that using the RD significantly improves model performance, achieving more than a 3\% increase in mean Average Precision for object detection with less than a 1\% increase in model parameters. Beyond 1-stage object detection models, the RD module improves the effectiveness of 2-stage models and DETR-based architectures, such as Faster R-CNN and Deformable DETR
본 보고서는 ECCV 2026과 함께 개최된 제8회 대규모 비디오 객체 분할(LSVOS) 챌린지의 모바일 통합 멀티모달 이해(MUMU) 트랙을 요약한다. MUMU는 0.5B 파라미터와 최대 추론 메모리 8GB의 예산 내에서, 품질, 장면, 이벤트 어휘에 대한 다중 개념 이미지 태깅, 개방 어휘 객체 탐지, 간결한 이미지 캡셔닝을 함께 수행하는 단일 통합 모델을 요구한다. 제출물은 효율 계수로 스케일된 가중 과제 점수로 순위를 매긴다. 본 보고서는 과제, 규칙, 평가 프로토콜을 설명하고, 최종 결과를 보고하며, 상위 세 팀의 방법을 검토한다. 선두 세 해법은 모두 소형 Florence-2-base 모델을 기반으로 하지만 서로 다른 방식으로 확장한다. 교차 과제 추론 정제를 갖춘 경량 태깅 헤드, 커버리지를 고려한 다중 소스 데이터로 학습하고 과제 의존적 디코딩을 사용하는 공유 저랭크 어댑터, 캡션을 태그로 매핑하고 탐지 테스트 시점 증강을 적용하는 학습 불필요 파이프라인이 그것이다. 결과는 감독, 디코딩, 교차 과제 단서가 각 과제에 맞게 설계될 때 소형 프롬프트 기반 비전-언어 모델이 세 가지 출력을 모두 담당할 수 있음을 보여주며, 새로운 범주에 대한 개방 어휘 탐지는 여전히 주요 병목으로 남아 있다.
Recently, Vision Transformer and its variants have shown great promise on various computer vision tasks. The ability of capturing short- and long-range visual dependencies through self-attention is arguably the main source for the success. But it also brings challenges due to quadratic computational overhead, especially for the high-resolution vision tasks (e.g., object detection). In this paper, we present focal self-attention, a new mechanism that incorporates both fine-grained local and coarse-grained global interactions. Using this new mechanism, each token attends the closest surrounding tokens at fine granularity but the tokens far away at coarse granularity, and thus can capture both short- and long-range visual dependencies efficiently and effectively. With focal self-attention, we propose a new variant of Vision Transformer models, called Focal Transformer, which achieves superior performance over the state-of-the-art vision Transformers on a range of public image classification and object detection benchmarks. In particular, our Focal Transformer models with a moderate size of 51.1M and a larger size of 89.8M achieve 83.5 and 83.8 Top-1 accuracy, respectively, on ImageNet classification at 224x224 resolution. Using Focal Transformers as the backbones, we obtain consistent and substantial improvements over the current state-of-the-art Swin Transformers for 6 different object detection methods trained with standard 1x and 3x schedules. Our largest Focal Transformer yields 58.7/58.9 box mAPs and 50.9/51.3 mask mAPs on COCO mini-val/test-dev, and 55.4 mIoU on ADE20K for semantic segmentation, creating new SoTA on three of the most challenging computer vision tasks.
In this work we study the impact of noise on the training of object detection networks for the medical domain, and how it can be mitigated by improving the training procedure. Annotating large medical datasets for training data-hungry deep learning models is expensive and time consuming. Leveraging information that is already collected in clinical practice, in the form of text reports, bookmarks or lesion measurements would substantially reduce this cost. Obtaining precise lesion bounding boxes through automatic mining procedures, however, is difficult. We provide here a quantitative evaluation of the effect of bounding box coordinate noise on the performance of Faster R-CNN object detection networks for breast mass detection. Varying degrees of noise are simulated by randomly modifying the bounding boxes: in our experiments, bounding boxes could be enlarged up to six times the original size. The noise is injected in the CBIS-DDSM collection, a well curated public mammography dataset for which accurate lesion location is available. We show how, due to an imperfect matching between the ground truth and the network bounding box proposals, the noise is propagated during training and reduces the ability of the network to correctly classify lesions from background. When using the standard Intersection over Union criterion, the area under the FROC curve decreases by up to 9%. A novel matching criterion is proposed to improve tolerance to noise.
Open-vocabulary detection (OVD) is an object detection task aiming at detecting objects from novel categories beyond the base categories on which the detector is trained. Recent OVD methods rely on large-scale visual-language pre-trained models, such as CLIP, for recognizing novel objects. We identify the two core obstacles that need to be tackled when incorporating these models into detector training: (1) the distribution mismatch that happens when applying a VL-model trained on whole images to region recognition tasks; (2) the difficulty of localizing objects of unseen classes. To overcome these obstacles, we propose CORA, a DETR-style framework that adapts CLIP for Open-vocabulary detection by Region prompting and Anchor pre-matching. Region prompting mitigates the whole-to-region distribution gap by prompting the region features of the CLIP-based region classifier. Anchor pre-matching helps learning generalizable object localization by a class-aware matching mechanism. We evaluate CORA on the COCO OVD benchmark, where we achieve 41.7 AP50 on novel classes, which outperforms the previous SOTA by 2.4 AP50 even without resorting to extra training data. When extra training data is available, we train CORA$^+$ on both ground-truth base-category annotations and additional pseudo bounding box labels computed by CORA. CORA$^+$ achieves 43.1 AP50 on the COCO OVD benchmark and 28.1 box APr on the LVIS OVD benchmark.
Multiple objects tracking finds its applications in many high level vision analysis like object behaviour interpretation and gait recognition. In this paper, a feature based method to track the multiple moving objects in surveillance video sequence is proposed. Object tracking is done by extracting the color and Hu moments features from the motion segmented object blob and establishing the association of objects in the successive frames of the video sequence based on Chi-Square dissimilarity measure and nearest neighbor classifier. The benchmark IEEE PETS and IEEE Change Detection datasets has been used to show the robustness of the proposed method. The proposed method is assessed quantitatively using the precision and recall accuracy metrics. Further, comparative evaluation with related works has been carried out to exhibit the efficacy of the proposed method.
Modern CNN-based object detectors rely on bounding box regression and non-maximum suppression to localize objects. While the probabilities for class labels naturally reflect classification confidence, localization confidence is absent. This makes properly localized bounding boxes degenerate during iterative regression or even suppressed during NMS. In the paper we propose IoU-Net learning to predict the IoU between each detected bounding box and the matched ground-truth. The network acquires this confidence of localization, which improves the NMS procedure by preserving accurately localized bounding boxes. Furthermore, an optimization-based bounding box refinement method is proposed, where the predicted IoU is formulated as the objective. Extensive experiments on the MS-COCO dataset show the effectiveness of IoU-Net, as well as its compatibility with and adaptivity to several state-of-the-art object detectors.
Sonar sensing is fundamental for underwater robotics, but limited by capabilities of AI systems, which need large training datasets. Public data in sonar modalities is lacking. This paper presents the Marine Debris Forward-Looking Sonar datasets, with three different settings (watertank, turntable, flooded quarry) increasing dataset diversity and multiple computer vision tasks: object classification, object detection, semantic segmentation, patch matching, and unsupervised learning. We provide full dataset description, basic analysis and initial results for some tasks. We expect the research community will benefit from this dataset, which is publicly available at https://doi.org/10.5281/zenodo.15101686
In saliency detection, every pixel needs contextual information to make saliency prediction. Previous models usually incorporate contexts holistically. However, for each pixel, usually only part of its context region is useful and contributes to its prediction, while some other part may serve as noises and distractions. In this paper, we propose a novel pixel-wise contextual attention network, \ie PiCANet, to learn to selectively attend to informative context locations at each pixel. Specifically, PiCANet generates an attention map over the context region of each pixel, where each attention weight corresponds to the relevance of a context location w.r.t the referred pixel. Then, attentive contextual features can be constructed via selectively incorporating the features of useful context locations with the learned attention. We propose three specific formulations of the PiCANet via embedding the pixel-wise contextual attention mechanism into the pooling and convolution operations with attending to global or local contexts. All the three models are fully differentiable and can be integrated with CNNs with joint training. We introduce the proposed PiCANets into a U-Net architecture for salient object detection. Experimental results indicate that the proposed PiCANets can significantly improve the saliency detection performance. The generated global and local attention can learn to incorporate global contrast and smoothness, respectively, which help localize salient objects more accurately and highlight them more uniformly. Consequently, our saliency model performs favorably against other state-of-the-art methods. Moreover, we also validate that PiCANets can also improve semantic segmentation and object detection performances, which further demonstrates their effectiveness and generalization ability.
Object detectors often suffer from the domain gap between training (source domain) and real-world applications (target domain). Mean-teacher self-training is a powerful paradigm in unsupervised domain adaptation for object detection, but it struggles with low-quality pseudo-labels. In this work, we identify the intriguing alignment and synergy between mean-teacher self-training and contrastive learning. Motivated by this, we propose Contrastive Mean Teacher (CMT) -- a unified, general-purpose framework with the two paradigms naturally integrated to maximize beneficial learning signals. Instead of using pseudo-labels solely for final predictions, our strategy extracts object-level features using pseudo-labels and optimizes them via contrastive learning, without requiring labels in the target domain. When combined with recent mean-teacher self-training methods, CMT leads to new state-of-the-art target-domain performance: 51.9% mAP on Foggy Cityscapes, outperforming the previously best by 2.1% mAP. Notably, CMT can stabilize performance and provide more significant gains as pseudo-label noise increases.
Creating an object detector, in computer vision, has some common challenges when initially developed based on Convolutional Neural Network (CNN) architecture. These challenges are more apparent when creating model that needs to adapt to images captured by various camera orientations, lighting conditions, and environmental changes. The availability of the initial training samples to cover all these conditions can be an enormous challenge with a time and cost burden. While the problem can exist when creating any type of object detection, some types are less common and have no pre-labeled image datasets that exists publicly. Sometime public datasets are not reliable nor comprehensive for a rare object type. Vehicle wheel is one of those example that been chosen to demonstrate the approach of creating a lighting and rotation invariant real-time detector based on YOLOv5 architecture. The objective is to provide a simple approach that could be used as a reference for developing other types of real-time object detectors.