We propose a novel end-to-end pipeline for online long-range vectorized high-definition (HD) map construction using on-board camera sensors. The vectorized representation of HD maps, employing polylines and polygons to represent map elements, is widely used by downstream tasks. However, previous schemes designed with reference to dynamic object detection overlook the structural constraints within linear map elements, resulting in performance degradation in long-range scenarios. In this paper, we exploit the properties of map elements to improve the performance of map construction. We extract more accurate bird's eye view (BEV) features guided by their linear structure, and then propose a hierarchical sparse map representation to further leverage the scalability of vectorized map elements and design a progressive decoding mechanism and a supervision strategy based on this representation. Our approach, ScalableMap, demonstrates superior performance on the nuScenes dataset, especially in long-range scenarios, surpassing previous state-of-the-art model by 6.5 mAP while achieving 18.3 FPS. Code is available at https://github.com/jingy1yu/ScalableMap.
Multi-view aggregation promises to overcome the occlusion and missed detection challenge in multi-object detection and tracking. Recent approaches in multi-view detection and 3D object detection made a huge performance leap by projecting all views to the ground plane and performing the detection in the Bird's Eye View (BEV). In this paper, we investigate if tracking in the BEV can also bring the next performance breakthrough in Multi-Target Multi-Camera (MTMC) tracking. Most current approaches in multi-view tracking perform the detection and tracking task in each view and use graph-based approaches to perform the association of the pedestrian across each view. This spatial association is already solved by detecting each pedestrian once in the BEV, leaving only the problem of temporal association. For the temporal association, we show how to learn strong Re-Identification (re-ID) features for each detection. The results show that early-fusion in the BEV achieves high accuracy for both detection and tracking. EarlyBird outperforms the state-of-the-art methods and improves the current state-of-the-art on Wildtrack by +4.6 MOTA and +5.6 IDF1.
A fundamental limitation of object detectors is that they suffer from "spatial bias", and in particular perform less satisfactorily when detecting objects near image borders. For a long time, there has been a lack of effective ways to measure and identify spatial bias, and little is known about where it comes from and what degree it is. To this end, we present a new zone evaluation protocol, extending from the traditional evaluation to a more generalized one, which measures the detection performance over zones, yielding a series of Zone Precisions (ZPs). For the first time, we provide numerical results, showing that the object detectors perform quite unevenly across the zones. Surprisingly, the detector's performance in the 96% border zone of the image does not reach the AP value (Average Precision, commonly regarded as the average detection performance in the entire image zone). To better understand spatial bias, a series of heuristic experiments are conducted. Our investigation excludes two intuitive conjectures about spatial bias that the object scale and the absolute positions of objects barely influence the spatial bias. We find that the key lies in the human-imperceptible divergence in data patterns between objects in different zones, thus eventually forming a visible performance gap between the zones. With these findings, we finally discuss a future direction for object detection, namely, spatial disequilibrium problem, aiming at pursuing a balanced detection ability over the entire image zone. By broadly evaluating 10 popular object detectors and 5 detection datasets, we shed light on the spatial bias of object detectors. We hope this work could raise a focus on detection robustness. The source codes, evaluation protocols, and tutorials are publicly available at https://github.com/Zzh-tju/ZoneEval.
Robotic crop phenotyping has emerged as a key technology to assess crops' morphological and physiological traits at scale. These phenotypical measurements are essential for developing new crop varieties with the aim of increasing productivity and dealing with environmental challenges such as climate change. However, developing and deploying crop phenotyping robots face many challenges such as complex and variable crop shapes that complicate robotic object detection, dynamic and unstructured environments that baffle robotic control, and real-time computing and managing big data that challenge robotic hardware/software. This work specifically tackles the first challenge by proposing a novel Digital-Twin(DT)MARS-CycleGAN model for image augmentation to improve our Modular Agricultural Robotic System (MARS)'s crop object detection from complex and variable backgrounds. Our core idea is that in addition to the cycle consistency losses in the CycleGAN model, we designed and enforced a new DT-MARS loss in the deep learning model to penalize the inconsistency between real crop images captured by MARS and synthesized images sensed by DT MARS. Therefore, the generated synthesized crop images closely mimic real images in terms of realism, and they are employed to fine-tune object detectors such as YOLOv8. Extensive experiments demonstrated that our new DT/MARS-CycleGAN framework significantly boosts our MARS' crop object/row detector's performance, contributing to the field of robotic crop phenotyping.
Source-free object detection (SFOD) aims to adapt a source-trained detector to an unlabeled target domain without access to the labeled source data. Current SFOD methods utilize a threshold-based pseudo-label approach in the adaptation phase, which is typically limited to high-confidence pseudo-labels and results in a loss of information. To address this issue, we propose a new approach to take full advantage of pseudo-labels by introducing high and low confidence thresholds. Specifically, the pseudo-labels with confidence scores above the high threshold are used conventionally, while those between the low and high thresholds are exploited using the Low-confidence Pseudo-labels Utilization (LPU) module. The LPU module consists of Proposal Soft Training (PST) and Local Spatial Contrastive Learning (LSCL). PST generates soft labels of proposals for soft training, which can mitigate the label mismatch problem. LSCL exploits the local spatial relationship of proposals to improve the model's ability to differentiate between spatially adjacent proposals, thereby optimizing representational features further. Combining the two components overcomes the challenges faced by traditional methods in utilizing low-confidence pseudo-labels. Extensive experiments on five cross-domain object detection benchmarks demonstrate that our proposed method outperforms the previous SFOD methods, achieving state-of-the-art performance.
Recent advancements in multimodal techniques open exciting possibilities for models excelling in diverse tasks involving text, audio, and image processing. Models like GPT-4V, blending computer vision and language modeling, excel in complex text and image tasks. Numerous prior research endeavors have diligently examined the performance of these Vision Large Language Models (VLLMs) across tasks like object detection, image captioning and others. However, these analyses often focus on evaluating the performance of each modality in isolation, lacking insights into their cross-modal interactions. Specifically, questions concerning whether these vision-language models execute vision and language tasks consistently or independently have remained unanswered. In this study, we draw inspiration from recent investigations into multilingualism and conduct a comprehensive analysis of model's cross-modal interactions. We introduce a systematic framework that quantifies the capability disparities between different modalities in the multi-modal setting and provide a set of datasets designed for these evaluations. Our findings reveal that models like GPT-4V tend to perform consistently modalities when the tasks are relatively simple. However, the trustworthiness of results derived from the vision modality diminishes as the tasks become more challenging. Expanding on our findings, we introduce "Vision Description Prompting," a method that effectively improves performance in challenging vision-related tasks.
While previous CNN-based models have exhibited promising results for salient object detection (SOD), their ability to explore global long-range dependencies is restricted. Our previous work, the Visual Saliency Transformer (VST), addressed this constraint from a transformer-based sequence-to-sequence perspective, to unify RGB and RGB-D SOD. In VST, we developed a multi-task transformer decoder that concurrently predicts saliency and boundary outcomes in a pure transformer architecture. Moreover, we introduced a novel token upsampling method called reverse T2T for predicting a high-resolution saliency map effortlessly within transformer-based structures. Building upon the VST model, we further propose an efficient and stronger VST version in this work, i.e. VST++. To mitigate the computational costs of the VST model, we propose a Select-Integrate Attention (SIA) module, partitioning foreground into fine-grained segments and aggregating background information into a single coarse-grained token. To incorporate 3D depth information with low cost, we design a novel depth position encoding method tailored for depth maps. Furthermore, we introduce a token-supervised prediction loss to provide straightforward guidance for the task-related tokens. We evaluate our VST++ model across various transformer-based backbones on RGB, RGB-D, and RGB-T SOD benchmark datasets. Experimental results show that our model outperforms existing methods while achieving a 25% reduction in computational costs without significant performance compromise. The demonstrated strong ability for generalization, enhanced performance, and heightened efficiency of our VST++ model highlight its potential.
Long-tailed object detection (LTOD) aims to handle the extreme data imbalance in real-world datasets, where many tail classes have scarce instances. One popular strategy is to explore extra data with image-level labels, yet it produces limited results due to (1) semantic ambiguity -- an image-level label only captures a salient part of the image, ignoring the remaining rich semantics within the image; and (2) location sensitivity -- the label highly depends on the locations and crops of the original image, which may change after data transformations like random cropping. To remedy this, we propose RichSem, a simple but effective method, which is robust to learn rich semantics from coarse locations without the need of accurate bounding boxes. RichSem leverages rich semantics from images, which are then served as additional soft supervision for training detectors. Specifically, we add a semantic branch to our detector to learn these soft semantics and enhance feature representations for long-tailed object detection. The semantic branch is only used for training and is removed during inference. RichSem achieves consistent improvements on both overall and rare-category of LVIS under different backbones and detectors. Our method achieves state-of-the-art performance without requiring complex training and testing procedures. Moreover, we show the effectiveness of our method on other long-tailed datasets with additional experiments. Code is available at \url{https://github.com/MengLcool/RichSem}.
Large Language Models (LLMs) have recently shown promise as high-level planners for robots when given access to a selection of low-level skills. However, it is often assumed that LLMs do not possess sufficient knowledge to be used for the low-level trajectories themselves. In this work, we address this assumption thoroughly, and investigate if an LLM (GPT-4) can directly predict a dense sequence of end-effector poses for manipulation tasks, when given access to only object detection and segmentation vision models. We designed a single, task-agnostic prompt, without any in-context examples, motion primitives, or external trajectory optimisers. Then we studied how well it can perform across 30 real-world language-based tasks, such as "open the bottle cap" and "wipe the plate with the sponge", and we investigated which design choices in this prompt are the most important. Our conclusions raise the assumed limit of LLMs for robotics, and we reveal for the first time that LLMs do indeed possess an understanding of low-level robot control sufficient for a range of common tasks, and that they can additionally detect failures and then re-plan trajectories accordingly. Videos, prompts, and code are available at: https://www.robot-learning.uk/language-models-trajectory-generators.
Automated event detection has emerged as one of the fundamental practices to monitor the behavior of technical systems by means of sensor data. In the automotive industry, these methods are in high demand for tracing events in time series data. For assessing the active vehicle safety systems, a diverse range of driving scenarios is conducted. These scenarios involve the recording of the vehicle's behavior using external sensors, enabling the evaluation of operational performance. In such setting, automated detection methods not only accelerate but also standardize and objectify the evaluation by avoiding subjective, human-based appraisals in the data inspection. This work proposes a novel event detection method that allows to identify frequency-based events in time series data. To this aim, the time series data is mapped to representations in the time-frequency domain, known as scalograms. After filtering scalograms to enhance relevant parts of the signal, an object detection model is trained to detect the desired event objects in the scalograms. For the analysis of unseen time series data, events can be detected in their scalograms with the trained object detection model and are thereafter mapped back to the time series data to mark the corresponding time interval. The algorithm, evaluated on unseen datasets, achieves a precision rate of 0.97 in event detection, providing sharp time interval boundaries whose accurate indication by human visual inspection is challenging. Incorporating this method into the vehicle development process enhances the accuracy and reliability of event detection, which holds major importance for rapid testing analysis.
Detecting objects in 3D space using multiple cameras, known as Multi-Camera 3D Object Detection (MC3D-Det), has gained prominence with the advent of bird's-eye view (BEV) approaches. However, these methods often struggle when faced with unfamiliar testing environments due to the lack of diverse training data encompassing various viewpoints and environments. To address this, we propose a novel method that aligns 3D detection with 2D camera plane results, ensuring consistent and accurate detections. Our framework, anchored in perspective debiasing, helps the learning of features resilient to domain shifts. In our approach, we render diverse view maps from BEV features and rectify the perspective bias of these maps, leveraging implicit foreground volumes to bridge the camera and BEV planes. This two-step process promotes the learning of perspective- and context-independent features, crucial for accurate object detection across varying viewpoints, camera parameters, and environmental conditions. Notably, our model-agnostic approach preserves the original network structure without incurring additional inference costs, facilitating seamless integration across various models and simplifying deployment. Furthermore, we also show our approach achieves satisfactory results in real data when trained only with virtual datasets, eliminating the need for real scene annotations. Experimental results on both Domain Generalization (DG) and Unsupervised Domain Adaptation (UDA) clearly demonstrate its effectiveness. The codes are available at https://github.com/EnVision-Research/Generalizable-BEV.
Intelligent transportation system combines advanced information technology to provide intelligent services such as monitoring, detection, and early warning for modern transportation. Intelligent transportation detection is the cornerstone of many intelligent traffic services by identifying task targets through object detection methods. However existing detection methods in intelligent transportation are limited by two aspects. First, there is a difference between the model knowledge pre-trained on large-scale datasets and the knowledge required for target task. Second, most detection models follow the pattern of single-source learning, which limits the learning ability. To address these problems, we propose a Multi Self-supervised Pre-fine-tuned Transformer Fusion (MSPTF) network, consisting of two steps: unsupervised pre-fine-tune domain knowledge learning and multi-model fusion target task learning. In the first step, we introduced self-supervised learning methods into transformer model pre-fine-tune which could reduce data costs and alleviate the knowledge gap between pre-trained model and target task. In the second step, we take feature information differences between different model architectures and different pre-fine-tune tasks into account and propose Multi-model Semantic Consistency Cross-attention Fusion (MSCCF) network to combine different transformer model features by considering channel semantic consistency and feature vector semantic consistency, which obtain more complete and proper fusion features for detection task. We experimented the proposed method on vehicle recognition dataset and road disease detection dataset and achieved 1.1%, 5.5%, 4.2% improvement compared with baseline and 0.7%, 1.8%, 1.7% compared with sota, which proved the effectiveness of our method.
Artificial intelligence has the potential to make valuable contributions to wildlife management through cost-effective methods for the collection and interpretation of wildlife data. Recent advances in remotely piloted aircraft systems (RPAS or ``drones'') and thermal imaging technology have created new approaches to collect wildlife data. These emerging technologies could provide promising alternatives to standard labourious field techniques as well as cover much larger areas. In this study, we conduct a comprehensive review and empirical study of drone-based wildlife detection. Specifically, we collect a realistic dataset of drone-derived wildlife thermal detections. Wildlife detections, including arboreal (for instance, koalas, phascolarctos cinereus) and ground dwelling species in our collected data are annotated via bounding boxes by experts. We then benchmark state-of-the-art object detection algorithms on our collected dataset. We use these experimental results to identify issues and discuss future directions in automatic animal monitoring using drones.
Monocular 3D object detection is an inherently ill-posed problem, as it is challenging to predict accurate 3D localization from a single image. Existing monocular 3D detection knowledge distillation methods usually project the LiDAR onto the image plane and train the teacher network accordingly. Transferring LiDAR-based model knowledge to RGB-based models is more complex, so a general distillation strategy is needed. To alleviate cross-modal prob-lem, we propose MonoSKD, a novel Knowledge Distillation framework for Monocular 3D detection based on Spearman correlation coefficient, to learn the relative correlation between cross-modal features. Considering the large gap between these features, strict alignment of features may mislead the training, so we propose a looser Spearman loss. Furthermore, by selecting appropriate distillation locations and removing redundant modules, our scheme saves more GPU resources and trains faster than existing methods. Extensive experiments are performed to verify the effectiveness of our framework on the challenging KITTI 3D object detection benchmark. Our method achieves state-of-the-art performance until submission with no additional inference computational cost. Our codes are available at https://github.com/Senwang98/MonoSKD
This paper focuses on a novel approach for detecting moving objects during camera motion. We present an optical-flow-based transformation that yields a consistent 2D invariant image output regardless of time instants, range of points in 3D, and the speed of the camera. In other words, this transformation generates a lookup image that remains invariant despite the changing projection of the 3D scene and camera motion. In the new domain, projections of 3D points that deviate from the values of the predefined lookup image can be clearly identified as moving relative to the stationary 3D environment, making them seamlessly detectable. The method does not require prior knowledge of the direction of motion or speed of the camera, nor does it necessitate 3D point range information. It is well-suited for real-time parallel processing, rendering it highly practical for implementation. We have validated the effectiveness of the new domain through simulations and experiments, demonstrating its robustness in scenarios involving rectilinear camera motion, both in simulations and with real-world data. This approach introduces new ways for moving objects detection during camera motion, and also lays the foundation for future research in the context of moving object detection during six-degrees-of-freedom camera motion.
Unsupervised salient object detection aims to detect salient objects without using supervision signals eliminating the tedious task of manually labeling salient objects. To improve training efficiency, end-to-end methods for USOD have been proposed as a promising alternative. However, current solutions rely heavily on noisy handcraft labels and fail to mine rich semantic information from deep features. In this paper, we propose a self-supervised end-to-end salient object detection framework via top-down context. Specifically, motivated by contrastive learning, we exploit the self-localization from the deepest feature to construct the location maps which are then leveraged to learn the most instructive segmentation guidance. Further considering the lack of detailed information in deepest features, we exploit the detail-boosting refiner module to enrich the location labels with details. Moreover, we observe that due to lack of supervision, current unsupervised saliency models tend to detect non-salient objects that are salient in some other samples of corresponding scenarios. To address this widespread issue, we design a novel Unsupervised Non-Salient Suppression (UNSS) method developing the ability to ignore non-salient objects. Extensive experiments on benchmark datasets demonstrate that our method achieves leading performance among the recent end-to-end methods and most of the multi-stage solutions. The code is available.
The flourishing success of Deep Neural Networks(DNNs) on RGB-input perception tasks has opened unbounded possibilities for non-RGB-input perception tasks, such as object detection from wireless signals, lidar scans, and infrared images. Compared to the matured development pipeline of RGB-input (source modality) models, developing non-RGB-input (target-modality) models from scratch poses excessive challenges in the modality-specific network design/training tricks and labor in the target-modality annotation. In this paper, we propose ModAlity Calibration (MAC), an efficient pipeline for calibrating target-modality inputs to the DNN object detection models developed on the RGB (source) modality. We compose a target-modality-input model by adding a small calibrator module ahead of a source-modality model and introduce MAC training techniques to impose dense supervision on the calibrator. By leveraging (1) prior knowledge synthesized from the source-modality model and (2) paired {target, source} data with zero manual annotations, our target-modality models reach comparable or better metrics than baseline models that require 100% manual annotations. We demonstrate the effectiveness of MAC by composing the WiFi-input, Lidar-input, and Thermal-Infrared-input models upon the pre-trained RGB-input models respectively.
Tracking microrobots is challenging, considering their minute size and high speed. As the field progresses towards developing microrobots for biomedical applications and conducting mechanistic studies in physiologically relevant media (e.g., collagen), this challenge is exacerbated by the dense surrounding environments with feature size and shape comparable to microrobots. Herein, we report Motion Enhanced Multi-level Tracker (MEMTrack), a robust pipeline for detecting and tracking microrobots using synthetic motion features, deep learning-based object detection, and a modified Simple Online and Real-time Tracking (SORT) algorithm with interpolation for tracking. Our object detection approach combines different models based on the object's motion pattern. We trained and validated our model using bacterial micro-motors in collagen (tissue phantom) and tested it in collagen and aqueous media. We demonstrate that MEMTrack accurately tracks even the most challenging bacteria missed by skilled human annotators, achieving precision and recall of 77% and 48% in collagen and 94% and 35% in liquid media, respectively. Moreover, we show that MEMTrack can quantify average bacteria speed with no statistically significant difference from the laboriously-produced manual tracking data. MEMTrack represents a significant contribution to microrobot localization and tracking, and opens the potential for vision-based deep learning approaches to microrobot control in dense and low-contrast settings. All source code for training and testing MEMTrack and reproducing the results of the paper have been made publicly available https://github.com/sawhney-medha/MEMTrack.
This study proposes a method based on lightweight convolutional neural networks (CNN) and generative adversarial networks (GAN) for apple ripeness and damage level detection tasks. Initially, a lightweight CNN model is designed by optimizing the model's depth and width, as well as employing advanced model compression techniques, successfully reducing the model's parameter and computational requirements, thus enhancing real-time performance in practical applications. Simultaneously, attention mechanisms are introduced, dynamically adjusting the importance of different feature layers to improve the performance in object detection tasks. To address the issues of sample imbalance and insufficient sample size, GANs are used to generate realistic apple images, expanding the training dataset and enhancing the model's recognition capability when faced with apples of varying ripeness and damage levels. Furthermore, by applying the object detection network for damage location annotation on damaged apples, the accuracy of damage level detection is improved, providing a more precise basis for decision-making. Experimental results show that in apple ripeness grading detection, the proposed model achieves 95.6\%, 93.8\%, 95.0\%, and 56.5 in precision, recall, accuracy, and FPS, respectively. In apple damage level detection, the proposed model reaches 95.3\%, 93.7\%, and 94.5\% in precision, recall, and mAP, respectively. In both tasks, the proposed method outperforms other mainstream models, demonstrating the excellent performance and high practical value of the proposed method in apple ripeness and damage level detection tasks.
We propose Re-parameterized Refocusing Convolution (RefConv) as a replacement for regular convolutional layers, which is a plug-and-play module to improve the performance without any inference costs. Specifically, given a pre-trained model, RefConv applies a trainable Refocusing Transformation to the basis kernels inherited from the pre-trained model to establish connections among the parameters. For example, a depth-wise RefConv can relate the parameters of a specific channel of convolution kernel to the parameters of the other kernel, i.e., make them refocus on the other parts of the model they have never attended to, rather than focus on the input features only. From another perspective, RefConv augments the priors of existing model structures by utilizing the representations encoded in the pre-trained parameters as the priors and refocusing on them to learn novel representations, thus further enhancing the representational capacity of the pre-trained model. Experimental results validated that RefConv can improve multiple CNN-based models by a clear margin on image classification (up to 1.47% higher top-1 accuracy on ImageNet), object detection and semantic segmentation without introducing any extra inference costs or altering the original model structure. Further studies demonstrated that RefConv can reduce the redundancy of channels and smooth the loss landscape, which explains its effectiveness.