Object detection and tracking is an essential perception task for enabling fully autonomous navigation in robotic systems. Edge robot systems such as small drones need to execute complex maneuvers at high-speeds with limited resources, which places strict constraints on the underlying algorithms and hardware. Traditionally, frame-based cameras are used for vision-based perception due to their rich spatial information and simplified synchronous sensing capabilities. However, obtaining detailed information across frames incurs high energy consumption and may not even be required. In addition, their low temporal resolution renders them ineffective in high-speed motion scenarios. Event-based cameras offer a biologically-inspired solution to this by capturing only changes in intensity levels at exceptionally high temporal resolution and low power consumption, making them ideal for high-speed motion scenarios. However, their asynchronous and sparse outputs are not natively suitable with conventional deep learning methods. In this work, we propose TOFFE, a lightweight hybrid framework for performing event-based object motion estimation (including pose, direction, and speed estimation), referred to as Object Flow. TOFFE integrates bio-inspired Spiking Neural Networks (SNNs) and conventional Analog Neural Networks (ANNs), to efficiently process events at high temporal resolutions while being simple to train. Additionally, we present a novel event-based synthetic dataset involving high-speed object motion to train TOFFE. Our experimental results show that TOFFE achieves 5.7x/8.3x reduction in energy consumption and 4.6x/5.8x reduction in latency on edge GPU(Jetson TX2)/hybrid hardware(Loihi-2 and Jetson TX2), compared to previous event-based object detection baselines.
The attribution of the author of an art piece is typically a laborious manual process, usually relying on subjective evaluations of expert figures. However, there are some situations in which quantitative features of the artwork can support these evaluations. The extraction of these features can sometimes be automated, for instance, with the use of Machine Learning (ML) techniques. An example of these features is represented by repeated, mechanically impressed patterns, called punches, present chiefly in 13th and 14th-century panel paintings from Tuscany. Previous research in art history showcased a strong connection between the shapes of punches and specific artists or workshops, suggesting the possibility of using these quantitative cues to support the attribution. In the present work, we first collect a dataset of large-scale images of these panel paintings. Then, using YOLOv10, a recent and popular object detection model, we train a ML pipeline to perform object detection on the punches contained in the images. Due to the large size of the images, the detection procedure is split across multiple frames by adopting a sliding-window approach with overlaps, after which the predictions are combined for the whole image using a custom non-maximal suppression routine. Our results indicate how art historians working in the field can reliably use our method for the identification and extraction of punches.
Supervised fine-tuning methods (SFT) perform great efficiency on artificial intelligence interpretation in SAR images, leveraging the powerful representation knowledge from pre-training models. Due to the lack of domain-specific pre-trained backbones in SAR images, the traditional strategies are loading the foundation pre-train models of natural scenes such as ImageNet, whose characteristics of images are extremely different from SAR images. This may hinder the model performance on downstream tasks when adopting SFT on small-scale annotated SAR data. In this paper, an self-supervised learning (SSL) method of masked image modeling based on Masked Auto-Encoders (MAE) is proposed to learn feature representations of SAR images during the pre-training process and benefit the object detection task in SAR images of SFT. The evaluation experiments on the large-scale SAR object detection benchmark named SARDet-100k verify that the proposed method captures proper latent representations of SAR images and improves the model generalization in downstream tasks by converting the pre-trained domain from natural scenes to SAR images through SSL. The proposed method achieves an improvement of 1.3 mAP on the SARDet-100k benchmark compared to only the SFT strategies.
This paper presents LiFT, a lightweight, fully quantized 3D object detection algorithm for LiDAR data, optimized for real-time inference on FPGA platforms. Through an in-depth analysis of FPGA-specific limitations, we identify a set of FPGA-induced constraints that shape the algorithm's design. These include a computational complexity limit of 30 GMACs (billion multiply-accumulate operations), INT8 quantization for weights and activations, 2D cell-based processing instead of 3D voxels, and minimal use of skip connections. To meet these constraints while maximizing performance, LiFT combines novel mechanisms with state-of-the-art techniques such as reparameterizable convolutions and fully sparse architecture. Key innovations include the Dual-bound Pillar Feature Net, which boosts performance without increasing complexity, and an efficient scheme for INT8 quantization of input features. With a computational cost of just 20.73 GMACs, LiFT stands out as one of the few algorithms targeting minimal-complexity 3D object detection. Among comparable methods, LiFT ranks first, achieving an mAP of 51.84% and an NDS of 61.01% on the challenging NuScenes validation dataset. The code will be available at https://github.com/vision-agh/lift.
Camouflaged object detection (COD) aims to distinguish hidden objects embedded in an environment highly similar to the object. Conventional video-based COD (VCOD) methods explicitly extract motion cues or employ complex deep learning networks to handle the temporal information, which is limited by high complexity and unstable performance. In this work, we propose a green VCOD method named GreenVCOD. Built upon a green ICOD method, GreenVCOD uses long- and short-term temporal neighborhoods (TN) to capture joint spatial/temporal context information for decision refinement. Experimental results show that GreenVCOD offers competitive performance compared to state-of-the-art VCOD benchmarks.
Advanced Driver Assistance Systems (ADAS) based on deep neural networks (DNNs) are widely used in autonomous vehicles for critical perception tasks such as object detection, semantic segmentation, and lane recognition. However, these systems are highly sensitive to input variations, such as noise and changes in lighting, which can compromise their effectiveness and potentially lead to safety-critical failures. This study offers a comprehensive empirical evaluation of image perturbations, techniques commonly used to assess the robustness of DNNs, to validate and improve the robustness and generalization of ADAS perception systems. We first conducted a systematic review of the literature, identifying 38 categories of perturbations. Next, we evaluated their effectiveness in revealing failures in two different ADAS, both at the component and at the system level. Finally, we explored the use of perturbation-based data augmentation and continuous learning strategies to improve ADAS adaptation to new operational design domains. Our results demonstrate that all categories of image perturbations successfully expose robustness issues in ADAS and that the use of dataset augmentation and continuous learning significantly improves ADAS performance in novel, unseen environments.
Low-light image enhancement (LLE) aims to improve the visual quality of images captured in poorly lit conditions, which often suffer from low brightness, low contrast, noise, and color distortions. These issues hinder the performance of computer vision tasks such as object detection, facial recognition, and autonomous driving.Traditional enhancement techniques, such as multi-scale fusion and histogram equalization, fail to preserve fine details and often struggle with maintaining the natural appearance of enhanced images under complex lighting conditions. Although the Retinex theory provides a foundation for image decomposition, it often amplifies noise, leading to suboptimal image quality. In this paper, we propose the Dual Light Enhance Network (DLEN), a novel architecture that incorporates two distinct attention mechanisms, considering both spatial and frequency domains. Our model introduces a learnable wavelet transform module in the illumination estimation phase, preserving high- and low-frequency components to enhance edge and texture details. Additionally, we design a dual-branch structure that leverages the power of the Transformer architecture to enhance both the illumination and structural components of the image.Through extensive experiments, our model outperforms state-of-the-art methods on standard benchmarks.Code is available here: https://github.com/LaLaLoXX/DLEN
With the advancement of Internet of Things (IoT) technology, underwater target detection and tracking have become increasingly important for ocean monitoring and resource management. Existing methods often fall short in handling high-noise and low-contrast images in complex underwater environments, lacking precision and robustness. This paper introduces a novel SVGS-DSGAT model that combines GraphSage, SVAM, and DSGAT modules, enhancing feature extraction and target detection capabilities through graph neural networks and attention mechanisms. The model integrates IoT technology to facilitate real-time data collection and processing, optimizing resource allocation and model responsiveness. Experimental results demonstrate that the SVGS-DSGAT model achieves an mAP of 40.8% on the URPC 2020 dataset and 41.5% on the SeaDronesSee dataset, significantly outperforming existing mainstream models. This IoT-enhanced approach not only excels in high-noise and complex backgrounds but also improves the overall efficiency and scalability of the system. This research provides an effective IoT solution for underwater target detection technology, offering significant practical application value and broad development prospects.
To mitigate the adverse effects of hard samples on the training of the Flying Bird Object Detection (FBOD) model for surveillance videos, we propose a Co-Paced Learning Based on Confidence (CPL-BC) strategy and apply this strategy to the training process of the FBOD model. This strategy involves maintaining two models with identical structures but different initial parameter configurations, which collaborate with each other to select easy samples with prediction confidence exceeding a set threshold for training. As training progresses, the strategy gradually lowers the threshold, allowing more samples to participate, enhancing the model's ability to recognize objects from easy to hard. Before applying the CPL-BC strategy to train the FBOD models, we initially trained the two FBOD models to equip them with the capability to assess the difficulty level of flying bird object samples. Experimental results on two different datasets of flying bird objects in surveillance videos demonstrate that, compared to other model learning strategies, CPL-BC significantly improves detection accuracy, verifying the effectiveness and advancement of this method.
Transformer-based methods have achieved remarkable performance in event-based object detection, owing to the global modeling ability. However, they neglect the influence of non-event and noisy regions and process them uniformly, leading to high computational overhead. To mitigate computation cost, some researchers propose window attention based sparsification strategies to discard unimportant regions, which sacrifices the global modeling ability and results in suboptimal performance. To achieve better trade-off between accuracy and efficiency, we propose Sparse Mamba (SMamba), which performs adaptive sparsification to reduce computational effort while maintaining global modeling capability. Specifically, a Spatio-Temporal Continuity Assessment module is proposed to measure the information content of tokens and discard uninformative ones by leveraging the spatiotemporal distribution differences between activity and noise events. Based on the assessment results, an Information-Prioritized Local Scan strategy is designed to shorten the scan distance between high-information tokens, facilitating interactions among them in the spatial dimension. Furthermore, to extend the global interaction from 2D space to 3D representations, a Global Channel Interaction module is proposed to aggregate channel information from a global spatial perspective. Results on three datasets (Gen1, 1Mpx, and eTram) demonstrate that our model outperforms other methods in both performance and efficiency.
The increasing availability of traffic videos functioning on a 24/7/365 time scale has the great potential of increasing the spatio-temporal coverage of traffic accidents, which will help improve traffic safety. However, analyzing footage from hundreds, if not thousands, of traffic cameras in a 24/7/365 working protocol remains an extremely challenging task, as current vision-based approaches primarily focus on extracting raw information, such as vehicle trajectories or individual object detection, but require laborious post-processing to derive actionable insights. We propose SeeUnsafe, a new framework that integrates Multimodal Large Language Model (MLLM) agents to transform video-based traffic accident analysis from a traditional extraction-then-explanation workflow to a more interactive, conversational approach. This shift significantly enhances processing throughput by automating complex tasks like video classification and visual grounding, while improving adaptability by enabling seamless adjustments to diverse traffic scenarios and user-defined queries. Our framework employs a severity-based aggregation strategy to handle videos of various lengths and a novel multimodal prompt to generate structured responses for review and evaluation and enable fine-grained visual grounding. We introduce IMS (Information Matching Score), a new MLLM-based metric for aligning structured responses with ground truth. We conduct extensive experiments on the Toyota Woven Traffic Safety dataset, demonstrating that SeeUnsafe effectively performs accident-aware video classification and visual grounding by leveraging off-the-shelf MLLMs. Source code will be available at \url{https://github.com/ai4ce/SeeUnsafe}.
Radar and LiDAR have been widely used in autonomous driving as LiDAR provides rich structure information, and radar demonstrates high robustness under adverse weather. Recent studies highlight the effectiveness of fusing radar and LiDAR point clouds. However, challenges remain due to the modality misalignment and information loss during feature extractions. To address these issues, we propose a 4D radar-LiDAR framework to mutually enhance their representations. Initially, the indicative features from radar are utilized to guide both radar and LiDAR geometric feature learning. Subsequently, to mitigate their sparsity gap, the shape information from LiDAR is used to enrich radar BEV features. Extensive experiments on the View-of-Delft (VoD) dataset demonstrate our approach's superiority over existing methods, achieving the highest mAP of 71.76% across the entire area and 86.36\% within the driving corridor. Especially for cars, we improve the AP by 4.17% and 4.20% due to the strong indicative features and symmetric shapes.
Source-free domain-adaptive object detection is an interesting but scarcely addressed topic. It aims at adapting a source-pretrained detector to a distinct target domain without resorting to source data during adaptation. So far, there is no data augmentation scheme tailored to source-free domain-adaptive object detection. To this end, this paper presents a novel data augmentation approach that cuts out target image regions where the detector is confident, augments them along with their respective pseudo-labels, and joins them into a challenging target image to adapt the detector. As the source data is out of reach during adaptation, we implement our approach within a teacher-student learning paradigm to ensure that the model does not collapse during the adaptation procedure. We evaluated our approach on three adaptation benchmarks of traffic scenes, scoring new state-of-the-art on two of them.
Current image tokenization methods require a large number of tokens to capture the information contained within images. Although the amount of information varies across images, most image tokenizers only support fixed-length tokenization, leading to inefficiency in token allocation. In this study, we introduce One-D-Piece, a discrete image tokenizer designed for variable-length tokenization, achieving quality-controllable mechanism. To enable variable compression rate, we introduce a simple but effective regularization mechanism named "Tail Token Drop" into discrete one-dimensional image tokenizers. This method encourages critical information to concentrate at the head of the token sequence, enabling support of variadic tokenization, while preserving state-of-the-art reconstruction quality. We evaluate our tokenizer across multiple reconstruction quality metrics and find that it delivers significantly better perceptual quality than existing quality-controllable compression methods, including JPEG and WebP, at smaller byte sizes. Furthermore, we assess our tokenizer on various downstream computer vision tasks, including image classification, object detection, semantic segmentation, and depth estimation, confirming its adaptability to numerous applications compared to other variable-rate methods. Our approach demonstrates the versatility of variable-length discrete image tokenization, establishing a new paradigm in both compression efficiency and reconstruction performance. Finally, we validate the effectiveness of tail token drop via detailed analysis of tokenizers.
Remote sensing (RS) visual tasks have gained significant academic and practical importance. However, they encounter numerous challenges that hinder effective feature extraction, including the detection and recognition of multiple objects exhibiting substantial variations in scale within a single image. While prior dual-branch or multi-branch architectural strategies have been effective in managing these object variances, they have concurrently resulted in considerable increases in computational demands and parameter counts. Consequently, these architectures are rendered less viable for deployment on resource-constrained devices. Contemporary lightweight backbone networks, designed primarily for natural images, frequently encounter difficulties in effectively extracting features from multi-scale objects, which compromises their efficacy in RS visual tasks. This article introduces LWGANet, a specialized lightweight backbone network tailored for RS visual tasks, incorporating a novel lightweight group attention (LWGA) module designed to address these specific challenges. LWGA module, tailored for RS imagery, adeptly harnesses redundant features to extract a wide range of spatial information, from local to global scales, without introducing additional complexity or computational overhead. This facilitates precise feature extraction across multiple scales within an efficient framework.LWGANet was rigorously evaluated across twelve datasets, which span four crucial RS visual tasks: scene classification, oriented object detection, semantic segmentation, and change detection. The results confirm LWGANet's widespread applicability and its ability to maintain an optimal balance between high performance and low complexity, achieving SOTA results across diverse datasets. LWGANet emerged as a novel solution for resource-limited scenarios requiring robust RS image processing capabilities.
Polysomnographic sleep analysis is the standard clinical method to accurately diagnose and treat sleep disorders. It is an intricate process which involves the manual identification, classification, and location of multiple sleep event patterns. This is complex, for which identification of different types of events involves focusing on different subsets of signals, resulting on an iterative time-consuming process entailing several visual analysis passes. In this paper we propose a multi-task deep-learning approach for the simultaneous detection of sleep events and hypnogram construction in one single pass. Taking as reference state-of-the-art methodology for object-detection in the field of Computer Vision, we reformulate the problem for the analysis of multi-variate time sequences, and more specifically for pattern detection in the sleep analysis scenario. We investigate the performance of the resulting method in identifying different assembly combinations of EEG arousals, respiratory events (apneas and hypopneas) and sleep stages, also considering different input signal montage configurations. Furthermore, we evaluate our approach using two independent datasets, assessing true-generalization effects involving local and external validation scenarios. Based on our results, we analyze and discuss our method's capabilities and its potential wide-range applicability across different settings and datasets.
LiDAR is a crucial sensor in autonomous driving, commonly used alongside cameras. By exploiting this camera-LiDAR setup and recent advances in image representation learning, prior studies have shown the promising potential of image-to-LiDAR distillation. These prior arts focus on the designs of their own losses to effectively distill the pre-trained 2D image representations into a 3D model. However, the other parts of the designs have been surprisingly unexplored. We find that fundamental design elements, e.g., the LiDAR coordinate system, quantization according to the existing input interface, and data utilization, are more critical than developing loss functions, which have been overlooked in prior works. In this work, we show that simple fixes to these designs notably outperform existing methods by 16% in 3D semantic segmentation on the nuScenes dataset and 13% in 3D object detection on the KITTI dataset in downstream task performance. We focus on overlooked design choices along the spatial and temporal axes. Spatially, prior work has used cylindrical coordinate and voxel sizes without considering their side effects yielded with a commonly deployed sparse convolution layer input interface, leading to spatial quantization errors in 3D models. Temporally, existing work has avoided cumbersome data curation by discarding unsynced data, limiting the use to only the small portion of data that is temporally synced across sensors. We analyze these effects and propose simple solutions for each overlooked aspect.
We explore the impact of aperture size and shape on automotive camera systems for deep-learning-based tasks like traffic sign recognition and light state detection. A method is proposed to simulate optical effects using the point spread function (PSF), enhancing realism and reducing the domain gap between synthetic and real-world images. Computer-generated scenes are refined with this technique to model optical distortions and improve simulation accuracy.
In soccer video analysis, player detection is essential for identifying key events and reconstructing tactical positions. The presence of numerous players and frequent occlusions, combined with copyright restrictions, severely restricts the availability of datasets, leaving limited options such as SoccerNet-Tracking and SportsMOT. These datasets suffer from a lack of diversity, which hinders algorithms from adapting effectively to varied soccer video contexts. To address these challenges, we developed SoccerSynth-Detection, the first synthetic dataset designed for the detection of synthetic soccer players. It includes a broad range of random lighting and textures, as well as simulated camera motion blur. We validated its efficacy using the object detection model (Yolov8n) against real-world datasets (SoccerNet-Tracking and SportsMoT). In transfer tests, it matched the performance of real datasets and significantly outperformed them in images with motion blur; in pre-training tests, it demonstrated its efficacy as a pre-training dataset, significantly enhancing the algorithm's overall performance. Our work demonstrates the potential of synthetic datasets to replace real datasets for algorithm training in the field of soccer video analysis.
Detecting the three-dimensional position and orientation of objects using a single RGB camera is a foundational task in computer vision with many important applications. Traditionally, 3D object detection methods are trained in a fully-supervised setup, requiring vast amounts of human annotations, which are laborious, costly, and do not scale well with the ever-increasing amounts of data being captured. In this paper, we present the first method to train 3D object detectors for monocular RGB cameras without domain-specific human annotations, thus making orders of magnitude more data available for training. Thanks to newly proposed Canonical Object Space, the method can not only exploit data across a variety of datasets and camera setups to train a single 3D detector, but unlike previous work it also works out of the box in previously unseen camera setups. All this is crucial for practical applications, where the data and cameras are extremely heterogeneous. The method is evaluated on two standard autonomous driving datasets, where it outperforms previous works, which, unlike our method, still rely on 2D human annotations.