Mobile robots are reaching unprecedented speeds, with platforms like Unitree B2, and Fraunhofer O3dyn achieving maximum speeds between 5 and 10 m/s. However, effectively utilizing such speeds remains a challenge due to the limitations of RGB cameras, which suffer from motion blur and fail to provide real-time responsiveness. Event cameras, with their asynchronous operation, and low-latency sensing, offer a promising alternative for high-speed robotic perception. In this work, we introduce MTevent, a dataset designed for 6D pose estimation and moving object detection in highly dynamic environments with large detection distances. Our setup consists of a stereo-event camera and an RGB camera, capturing 75 scenes, each on average 16 seconds, and featuring 16 unique objects under challenging conditions such as extreme viewing angles, varying lighting, and occlusions. MTevent is the first dataset to combine high-speed motion, long-range perception, and real-world object interactions, making it a valuable resource for advancing event-based vision in robotics. To establish a baseline, we evaluate the task of 6D pose estimation using NVIDIA's FoundationPose on RGB images, achieving an Average Recall of 0.22 with ground-truth masks, highlighting the limitations of RGB-based approaches in such dynamic settings. With MTevent, we provide a novel resource to improve perception models and foster further research in high-speed robotic vision. The dataset is available for download https://huggingface.co/datasets/anas-gouda/MTevent
Accurate and fast extraction of foreground object is a key prerequisite for a wide range of computer vision applications such as object tracking and recognition. Thus, enormous background subtraction methods for foreground object detection have been proposed in recent decades. However, it is still regarded as a tough problem due to a variety of challenges such as illumination variations, camera jitter, dynamic backgrounds, shadows, and so on. Currently, there is no single method that can handle all the challenges in a robust way. In this letter, we try to solve this problem from a new perspective by combining different state-of-the-art background subtraction algorithms to create a more robust and more advanced foreground detection algorithm. More specifically, an encoder-decoder fully convolutional neural network architecture is trained to automatically learn how to leverage the characteristics of different algorithms to fuse the results produced by different background subtraction algorithms and output a more precise result. Comprehensive experiments evaluated on the CDnet 2014 dataset demonstrate that the proposed method outperforms all the considered single background subtraction algorithm. And we show that our solution is more efficient than other combination strategies.
The objective of this paper is few-shot object detection (FSOD) -- the task of expanding an object detector for a new category given only a few instances for training. We introduce a simple pseudo-labelling method to source high-quality pseudo-annotations from the training set, for each new category, vastly increasing the number of training instances and reducing class imbalance; our method finds previously unlabelled instances. Na\"ively training with model predictions yields sub-optimal performance; we present two novel methods to improve the precision of the pseudo-labelling process: first, we introduce a verification technique to remove candidate detections with incorrect class labels; second, we train a specialised model to correct poor quality bounding boxes. After these two novel steps, we obtain a large set of high-quality pseudo-annotations that allow our final detector to be trained end-to-end. Additionally, we demonstrate our method maintains base class performance, and the utility of simple augmentations in FSOD. While benchmarking on PASCAL VOC and MS-COCO, our method achieves state-of-the-art or second-best performance compared to existing approaches across all number of shots.
The field of object detection has made significant advances riding on the wave of region-based ConvNets, but their training procedure still includes many heuristics and hyperparameters that are costly to tune. We present a simple yet surprisingly effective online hard example mining (OHEM) algorithm for training region-based ConvNet detectors. Our motivation is the same as it has always been -- detection datasets contain an overwhelming number of easy examples and a small number of hard examples. Automatic selection of these hard examples can make training more effective and efficient. OHEM is a simple and intuitive algorithm that eliminates several heuristics and hyperparameters in common use. But more importantly, it yields consistent and significant boosts in detection performance on benchmarks like PASCAL VOC 2007 and 2012. Its effectiveness increases as datasets become larger and more difficult, as demonstrated by the results on the MS COCO dataset. Moreover, combined with complementary advances in the field, OHEM leads to state-of-the-art results of 78.9% and 76.3% mAP on PASCAL VOC 2007 and 2012 respectively.
This study delves into semi-supervised object detection (SSOD) to improve detector performance with additional unlabeled data. State-of-the-art SSOD performance has been achieved recently by self-training, in which training supervision consists of ground truths and pseudo-labels. In current studies, we observe that class imbalance in SSOD severely impedes the effectiveness of self-training. To address the class imbalance, we propose adaptive class-rebalancing self-training (ACRST) with a novel memory module called CropBank. ACRST adaptively rebalances the training data with foreground instances extracted from the CropBank, thereby alleviating the class imbalance. Owing to the high complexity of detection tasks, we observe that both self-training and data-rebalancing suffer from noisy pseudo-labels in SSOD. Therefore, we propose a novel two-stage filtering algorithm to generate accurate pseudo-labels. Our method achieves satisfactory improvements on MS-COCO and VOC benchmarks. When using only 1\% labeled data in MS-COCO, our method achieves 17.02 mAP improvement over supervised baselines, and 5.32 mAP improvement compared with state-of-the-art methods.
Recently significant progress has been made in pedestrian detection, but it remains challenging to achieve high performance in occluded and crowded scenes. It could be attributed mostly to the widely used representation of pedestrians, i.e., 2D axis-aligned bounding box, which just describes the approximate location and size of the object. Bounding box models the object as a uniform distribution within the boundary, making pedestrians indistinguishable in occluded and crowded scenes due to much noise. To eliminate the problem, we propose a novel representation based on 2D beta distribution, named Beta Representation. It pictures a pedestrian by explicitly constructing the relationship between full-body and visible boxes, and emphasizes the center of visual mass by assigning different probability values to pixels. As a result, Beta Representation is much better for distinguishing highly-overlapped instances in crowded scenes with a new NMS strategy named BetaNMS. What's more, to fully exploit Beta Representation, a novel pipeline Beta R-CNN equipped with BetaHead and BetaMask is proposed, leading to high detection performance in occluded and crowded scenes.
Small oriented objects that represent tiny pixel-area in large-scale aerial images are difficult to detect due to their size and orientation. Existing oriented aerial detectors have shown promising results but are mainly focused on orientation modeling with less regard to the size of the objects. In this work, we proposed a method to accurately detect small oriented objects in aerial images by enhancing the classification and regression tasks of the oriented object detection model. We designed the Attention-Points Network consisting of two losses: Guided-Attention Loss (GALoss) and Box-Points Loss (BPLoss). GALoss uses an instance segmentation mask as ground-truth to learn the attention features needed to improve the detection of small objects. These attention features are then used to predict box points for BPLoss, which determines the points' position relative to the target oriented bounding box. Experimental results show the effectiveness of our Attention-Points Network on a standard oriented aerial dataset with small object instances (DOTA-v1.5) and on a maritime-related dataset (HRSC2016). The code is publicly available.
Efforts to connect LiDAR data with text, such as LidarCLIP, have primarily focused on embedding 3D point clouds into CLIP text-image space. However, these approaches rely on 3D point clouds, which present challenges in encoding efficiency and neural network processing. With the advent of advanced LiDAR sensors like Ouster OS1, which, in addition to 3D point clouds, produce fixed resolution depth, signal, and ambient panoramic 2D images, new opportunities emerge for LiDAR based tasks. In this work, we propose an alternative approach to connect LiDAR data with text by leveraging 2D imagery generated by the OS1 sensor instead of 3D point clouds. Using the Florence 2 large model in a zero-shot setting, we perform image captioning and object detection. Our experiments demonstrate that Florence 2 generates more informative captions and achieves superior performance in object detection tasks compared to existing methods like CLIP. By combining advanced LiDAR sensor data with a large pre-trained model, our approach provides a robust and accurate solution for challenging detection scenarios, including real-time applications requiring high accuracy and robustness.
The goal of this paper is to detect objects by exploiting their interrelationships. Contrary to existing methods, which learn objects and relations separately, our key idea is to learn the object-relation distribution jointly. We first propose a novel way of creating a graphical representation of an image from inter-object relation priors and initial class predictions, we call a context-likelihood graph. We then learn the joint distribution with an energy-based modeling technique which allows to sample and refine the context-likelihood graph iteratively for a given image. Our formulation of jointly learning the distribution enables us to generate a more accurate graph representation of an image which leads to a better object detection performance. We demonstrate the benefits of our context-likelihood graph formulation and the energy-based graph refinement via experiments on the Visual Genome and MS-COCO datasets where we achieve a consistent improvement over object detectors like DETR and Faster-RCNN, as well as alternative methods modeling object interrelationships separately. Our method is detector agnostic, end-to-end trainable, and especially beneficial for rare object classes.
Multiple-instance learning (MIL) has served as an important tool for a wide range of vision applications, for instance, image classification, object detection, and visual tracking. In this paper, we propose a novel method to solve the classical MIL problem, named relaxed multiple-instance SVM (RMI-SVM). We treat the positiveness of instance as a continuous variable, use Noisy-OR model to enforce the MIL constraints, and jointly optimize the bag label and instance label in a unified framework. The optimization problem can be efficiently solved using stochastic gradient decent. The extensive experiments demonstrate that RMI-SVM consistently achieves superior performance on various benchmarks for MIL. Moreover, we simply applied RMI-SVM to a challenging vision task, common object discovery. The state-of-the-art results of object discovery on Pascal VOC datasets further confirm the advantages of the proposed method.
Object recognition in video is an important task for plenty of applications, including autonomous driving perception, surveillance tasks, wearable devices or IoT networks. Object recognition using video data is more challenging than using still images due to blur, occlusions or rare object poses. Specific video detectors with high computational cost or standard image detectors together with a fast post-processing algorithm achieve the current state-of-the-art. This work introduces a novel post-processing pipeline that overcomes some of the limitations of previous post-processing methods by introducing a learning-based similarity evaluation between detections across frames. Our method improves the results of state-of-the-art specific video detectors, specially regarding fast moving objects, and presents low resource requirements. And applied to efficient still image detectors, such as YOLO, provides comparable results to much more computationally intensive detectors.
We present a novel one-shot method for object detection and 6 DoF pose estimation, that does not require training on target objects. At test time, it takes as input a target image and a textured 3D query model. The core idea is to represent a 3D model with a number of 2D templates rendered from different viewpoints. This enables CNN-based direct dense feature extraction and matching. The object is first localized in 2D, then its approximate viewpoint is estimated, followed by dense 2D-3D correspondence prediction. The final pose is computed with PnP. We evaluate the method on LineMOD, Occlusion, Homebrewed, YCB-V and TLESS datasets and report very competitive performance in comparison to the state-of-the-art methods trained on synthetic data, even though our method is not trained on the object models used for testing.
Object Referring Analysis (ORA), commonly known as referring expression comprehension, requires the identification and localization of specific objects in an image based on natural descriptions. Unlike generic object detection, ORA requires both accurate language understanding and precise visual localization, making it inherently more complex. Although recent pre-trained large visual grounding detectors have achieved significant progress, they heavily rely on extensively labeled data and time-consuming learning. To address these, we introduce a novel, training-free framework for zero-shot ORA, termed FLORA (Formal Language for Object Referring and Analysis). FLORA harnesses the inherent reasoning capabilities of large language models (LLMs) and integrates a formal language model - a logical framework that regulates language within structured, rule-based descriptions - to provide effective zero-shot ORA. More specifically, our formal language model (FLM) enables an effective, logic-driven interpretation of object descriptions without necessitating any training processes. Built upon FLM-regulated LLM outputs, we further devise a Bayesian inference framework and employ appropriate off-the-shelf interpretive models to finalize the reasoning, delivering favorable robustness against LLM hallucinations and compelling ORA performance in a training-free manner. In practice, our FLORA boosts the zero-shot performance of existing pretrained grounding detectors by up to around 45%. Our comprehensive evaluation across different challenging datasets also confirms that FLORA consistently surpasses current state-of-the-art zero-shot methods in both detection and segmentation tasks associated with zero-shot ORA. We believe our probabilistic parsing and reasoning of the LLM outputs elevate the reliability and interpretability of zero-shot ORA. We shall release codes upon publication.
LiDAR 3D object detection models are inevitably biased towards their training dataset. The detector clearly exhibits this bias when employed on a target dataset, particularly towards object sizes. However, object sizes vary heavily between domains due to, for instance, different labeling policies or geographical locations. State-of-the-art unsupervised domain adaptation approaches outsource methods to overcome the object size bias. Mainstream size adaptation approaches exploit target domain statistics, contradicting the original unsupervised assumption. Our novel unsupervised anchor calibration method addresses this limitation. Given a model trained on the source data, we estimate the optimal target anchors in a completely unsupervised manner. The main idea stems from an intuitive observation: by varying the anchor sizes for the target domain, we inevitably introduce noise or even remove valuable object cues. The latent object representation, perturbed by the anchor size, is closest to the learned source features only under the optimal target anchors. We leverage this observation for anchor size optimization. Our experimental results show that, without any retraining, we achieve competitive results even compared to state-of-the-art weakly-supervised size adaptation approaches. In addition, our anchor calibration can be combined with such existing methods, making them completely unsupervised.
6D object pose estimation is the problem of identifying the position and orientation of an object relative to a chosen coordinate system, which is a core technology for modern XR applications. State-of-the-art 6D object pose estimators directly predict an object pose given an object observation. Due to the ill-posed nature of the pose estimation problem, where multiple different poses can correspond to a single observation, generating additional plausible estimates per observation can be valuable. To address this, we reformulate the state-of-the-art algorithm GDRNPP and introduce EPRO-GDR (End-to-End Probabilistic Geometry-Guided Regression). Instead of predicting a single pose per detection, we estimate a probability density distribution of the pose. Using the evaluation procedure defined by the BOP (Benchmark for 6D Object Pose Estimation) Challenge, we test our approach on four of its core datasets and demonstrate superior quantitative results for EPRO-GDR on LM-O, YCB-V, and ITODD. Our probabilistic solution shows that predicting a pose distribution instead of a single pose can improve state-of-the-art single-view pose estimation while providing the additional benefit of being able to sample multiple meaningful pose candidates.
Recently, end-to-end object detectors have gained significant attention from the research community due to their outstanding performance. However, DETR typically relies on supervised pretraining of the backbone on ImageNet, which limits the practical application of DETR and the design of the backbone, affecting the model's potential generalization ability. In this paper, we propose a new training method called step-by-step training. Specifically, in the first stage, the one-to-many pre-trained YOLO detector is used to initialize the end-to-end detector. In the second stage, the backbone and encoder are consistent with the DETR-like model, but only the detector needs to be trained from scratch. Due to this training method, the object detector does not need the additional dataset (ImageNet) to train the backbone, which makes the design of the backbone more flexible and dramatically reduces the training cost of the detector, which is helpful for the practical application of the object detector. At the same time, compared with the DETR-like model, the step-by-step training method can achieve higher accuracy than the traditional training method of the DETR-like model. With the aid of this novel training method, we propose a brand-new end-to-end real-time object detection model called DEYOv3. DEYOv3-N achieves 41.1% on COCO val2017 and 270 FPS on T4 GPU, while DEYOv3-L achieves 51.3% AP and 102 FPS. Without the use of additional training data, DEYOv3 surpasses all existing real-time object detectors in terms of both speed and accuracy. It is worth noting that for models of N, S, and M scales, the training on the COCO dataset can be completed using a single 24GB RTX3090 GPU. Code will be released at https://github.com/ouyanghaodong/DEYOv3.
신뢰할 수 있는 교통 표지판 검출은 자율주행 시스템의 전 세계적 배포를 위한 전제 조건이며, 규제 준수와 도로 안전은 지역·거리·기상 조건 전반에 걸쳐 표지판을 올바르게 인지하는 데 달려 있다. 최근 진전에도 불구하고, 비전 기반 방법은 여전히 세 가지 근본적 한계에 직면한다. 국가 간 높은 다양성으로 인한 낮은 지역 간 일반화, 원거리에서의 소형 객체 검출 성능 저하(200m에서 교통 표지판이 $10{\times}10$ 픽셀에 불과할 수 있음), 차량이 표지판에 접근할 때 발생하는 강한 비선형 원근 왜곡 하에서의 취약한 시간적 추적이다. 본 논문에서는 카메라와 Light Detection and Ranging(LiDAR) 감지를 결합하여 견고하고 원거리이며 지역에 구애받지 않는 교통 표지판 인지 문제를 다룬다. Intensity-Aware Deformable Fusion 모듈을 갖춘 멀티모달 검출 프레임워크를 제안하며, 이 모듈은 후방반사 LiDAR 단서를 카메라 특징과 정렬하여 지역별 시각적 외관이 아닌 기하 불변량에 검출을 고정한다. 또한 차량 접근 시 비선형 원근 변환을 명시적으로 고려하는 이중 운동 모델 추적기를 도입하여, 선형 운동 가정 대비 시간적 일관성을 크게 향상시킨다. 더하여 가림(occlusion) 수준, 가독성(readability), 표지판 내장성(sign embeddedness), 도로 관련성(road relevance)을 추정하는 의미 속성 분류 파이프라인을 개발하여 하위 계획(planning)에 실행 가능한 맥락을 제공한다. 60개 이상 국가와 2,500시간 이상의 주행 데이터를 포괄하는 자체 데이터셋에 대한 광범위한 평가에서, 제안 파이프라인은 221,068개 평가 시퀀스 전반에 걸쳐 Object Miss Ratio(OMR) 0.49%를 달성하여, 상용급 자율주행 시스템에서 전 세계적으로 일반화 가능한 교통 표지판 인지를 입증한다.
산업 현장의 이상 탐지(anomaly detection)는 서로 다른 두 가지 결함 유형을 다루어야 한다. 하나는 국소적인 텍스처 손상으로 나타나는 구조적 이상(structural anomaly)이고, 다른 하나는 객체의 수, 구성 또는 배치에 관한 전역 규칙을 위반하는 논리적 이상(logical anomaly)이다. 기존 탐지기는 대체로 한 유형을 우선하는 대신 다른 유형을 희생한다. 특히 학습이 필요 없는(training-free) 방법은 고정된(frozen) 표현을 효과적으로 활용하지만 객체 수에 대한 명시적인 개념이 없으며, 객체 수를 추론하는 방법은 대개 범주별 구성 요소 모델링에 의존한다. 우리는 추가 학습이나 부품 수준의 지도(supervision) 없이도 학습이 필요 없는 이상 탐지에 카운팅 능력을 도입할 수 있음을 보인다. 핵심 아이디어는 정상 이미지의 통계를 이용해 이질적인 이상 단서들을 정렬하는 정상 집합 보정(normal-set calibration)으로, 이를 통해 통합된 학습 불필요 프레임워크 안에서 단서들을 직접 융합할 수 있다. 이 보정을 기반으로 하는 우리의 탐지기는 상호 보완적인 고정 단서들을 결합하여 논리적 이상과 구조적 이상을 모두 다룬다. MVTec-LOCO에서 우리 방법은 논리적 이상과 구조적 이상에 대해 각각 89.0과 95.9의 이미지 수준 AUROC를 달성하여 평균 92.5를 기록했으며, 이는 비교 대상 중 학습이 필요 없는 탐지기 가운데 가장 높은 값이다. 또한 네트워크 학습이나 부품 주석(annotation)을 요구하는 방법들과도 경쟁력을 유지하며, 구조적 변형 모델은 MVTec-AD에서 PatchCore와 동등한 성능(이미지 AUROC 99.1)을 보여, 제안한 보정이 논리적 이상 탐지를 넘어 일반화될 수 있음을 시사한다.
Object detection on drone-captured scenarios is a recent popular task. As drones always navigate in different altitudes, the object scale varies violently, which burdens the optimization of networks. Moreover, high-speed and low-altitude flight bring in the motion blur on the densely packed objects, which leads to great challenge of object distinction. To solve the two issues mentioned above, we propose TPH-YOLOv5. Based on YOLOv5, we add one more prediction head to detect different-scale objects. Then we replace the original prediction heads with Transformer Prediction Heads (TPH) to explore the prediction potential with self-attention mechanism. We also integrate convolutional block attention model (CBAM) to find attention region on scenarios with dense objects. To achieve more improvement of our proposed TPH-YOLOv5, we provide bags of useful strategies such as data augmentation, multiscale testing, multi-model integration and utilizing extra classifier. Extensive experiments on dataset VisDrone2021 show that TPH-YOLOv5 have good performance with impressive interpretability on drone-captured scenarios. On DET-test-challenge dataset, the AP result of TPH-YOLOv5 are 39.18%, which is better than previous SOTA method (DPNetV3) by 1.81%. On VisDrone Challenge 2021, TPHYOLOv5 wins 5th place and achieves well-matched results with 1st place model (AP 39.43%). Compared to baseline model (YOLOv5), TPH-YOLOv5 improves about 7%, which is encouraging and competitive.
In recent years, numerous domain adaptive strategies have been proposed to help deep learning models overcome the challenges posed by domain shift. However, even unsupervised domain adaptive strategies still require a large amount of target data. Medical imaging datasets are often characterized by class imbalance and scarcity of labeled and unlabeled data. Few-shot domain adaptive object detection (FSDAOD) addresses the challenge of adapting object detectors to target domains with limited labeled data. Existing works struggle with randomly selected target domain images that may not accurately represent the real population, resulting in overfitting to small validation sets and poor generalization to larger test sets. Medical datasets exhibit high class imbalance and background similarity, leading to increased false positives and lower mean Average Precision (map) in target domains. To overcome these challenges, we propose a novel FSDAOD strategy for microscopic imaging. Our contributions include a domain adaptive class balancing strategy for few-shot scenarios, multi-layer instance-level inter and intra-domain alignment to enhance similarity between class instances regardless of domain, and an instance-level classification loss applied in the middle layers of the object detector to enforce feature retention necessary for correct classification across domains. Extensive experimental results with competitive baselines demonstrate the effectiveness of our approach, achieving state-of-the-art results on two public microscopic datasets. Code available at https://github.co/intelligentMachinesLab/few-shot-domain-adaptive-microscopy