Existing studies on salient object detection (SOD) focus on extracting distinct objects with edge information and aggregating multi-level features to improve SOD performance. To achieve satisfactory performance, the methods employ refined edge information and low multi-level discrepancy. However, both performance gain and computational efficiency cannot be attained, which has motivated us to study the inefficiencies in existing encoder-decoder structures to avoid this trade-off. We propose TRACER, which detects salient objects with explicit edges by incorporating attention guided tracing modules. We employ a masked edge attention module at the end of the first encoder using a fast Fourier transform to propagate the refined edge information to the downstream feature extraction. In the multi-level aggregation phase, the union attention module identifies the complementary channel and important spatial information. To improve the decoder performance and computational efficiency, we minimize the decoder block usage with object attention module. This module extracts undetected objects and edge information from refined channels and spatial representations. Subsequently, we propose an adaptive pixel intensity loss function to deal with the relatively important pixels unlike conventional loss functions which treat all pixels equally. A comparison with 13 existing methods reveals that TRACER achieves state-of-the-art performance on five benchmark datasets. We have released TRACER at https://github.com/Karel911/TRACER.
Synthetic data has been a critical tool for training scene text detection and recognition models. On the one hand, synthetic word images have proven to be a successful substitute for real images in training scene text recognizers. On the other hand, however, scene text detectors still heavily rely on a large amount of manually annotated real-world images, which are expensive. In this paper, we introduce UnrealText, an efficient image synthesis method that renders realistic images via a 3D graphics engine. 3D synthetic engine provides realistic appearance by rendering scene and text as a whole, and allows for better text region proposals with access to precise scene information, e.g. normal and even object meshes. The comprehensive experiments verify its effectiveness on both scene text detection and recognition. We also generate a multilingual version for future research into multilingual scene text detection and recognition. Additionally, we re-annotate scene text recognition datasets in a case-sensitive way and include punctuation marks for more comprehensive evaluations. The code and the generated datasets are released at: https://github.com/Jyouhou/UnrealText/ .
Salient object detection (SOD) in optical remote sensing images (RSIs), or RSI-SOD, is an emerging topic in understanding optical RSIs. However, due to the difference between optical RSIs and natural scene images (NSIs), directly applying NSI-SOD methods to optical RSIs fails to achieve satisfactory results. In this paper, we propose a novel Adjacent Context Coordination Network (ACCoNet) to explore the coordination of adjacent features in an encoder-decoder architecture for RSI-SOD. Specifically, ACCoNet consists of three parts: an encoder, Adjacent Context Coordination Modules (ACCoMs), and a decoder. As the key component of ACCoNet, ACCoM activates the salient regions of output features of the encoder and transmits them to the decoder. ACCoM contains a local branch and two adjacent branches to coordinate the multi-level features simultaneously. The local branch highlights the salient regions in an adaptive way, while the adjacent branches introduce global information of adjacent levels to enhance salient regions. Additionally, to extend the capabilities of the classic decoder block (i.e., several cascaded convolutional layers), we extend it with two bifurcations and propose a Bifurcation-Aggregation Block to capture the contextual information in the decoder. Extensive experiments on two benchmark datasets demonstrate that the proposed ACCoNet outperforms 22 state-of-the-art methods under nine evaluation metrics, and runs up to 81 fps on a single NVIDIA Titan X GPU. The code and results of our method are available at https://github.com/MathLee/ACCoNet.
Millimeter-wave radars are being increasingly integrated into commercial vehicles to support new advanced driver-assistance systems by enabling robust and high-performance object detection, localization, as well as recognition - a key component of new environmental perception. In this paper, we propose a novel radar multiple-perspectives convolutional neural network (RAMP-CNN) that extracts the location and class of objects based on further processing of the range-velocity-angle (RVA) heatmap sequences. To bypass the complexity of 4D convolutional neural networks (NN), we propose to combine several lower-dimension NN models within our RAMP-CNN model that nonetheless approaches the performance upper-bound with lower complexity. The extensive experiments show that the proposed RAMP-CNN model achieves better average recall and average precision than prior works in all testing scenarios. Besides, the RAMP-CNN model is validated to work robustly under nighttime, which enables low-cost radars as a potential substitute for pure optical sensing under severe conditions.
To improve instance-level detection/segmentation performance, existing self-supervised and semi-supervised methods extract either task-unrelated or task-specific training signals from unlabeled data. We show that these two approaches, at the two extreme ends of the task-specificity spectrum, are suboptimal for the task performance. Utilizing too little task-specific training signals causes underfitting to the ground-truth labels of downstream tasks, while the opposite causes overfitting to the ground-truth labels. To this end, we propose a novel Class-Agnostic Semi-Supervised Learning (CA-SSL) framework to achieve a more favorable task-specificity balance in extracting training signals from unlabeled data. CA-SSL has three training stages that act on either ground-truth labels (labeled data) or pseudo labels (unlabeled data). This decoupling strategy avoids the complicated scheme in traditional SSL methods that balances the contributions from both data types. Especially, we introduce a warmup training stage to achieve a more optimal balance in task specificity by ignoring class information in the pseudo labels, while preserving localization training signals. As a result, our warmup model can better avoid underfitting/overfitting when fine-tuned on the ground-truth labels in detection and segmentation tasks. Using 3.6M unlabeled data, we achieve a significant performance gain of 4.7% over ImageNet-pretrained baseline on FCOS object detection. In addition, our warmup model demonstrates excellent transferability to other detection and segmentation frameworks.
In this paper, we propose a novel training strategy called SupFusion, which provides an auxiliary feature level supervision for effective LiDAR-Camera fusion and significantly boosts detection performance. Our strategy involves a data enhancement method named Polar Sampling, which densifies sparse objects and trains an assistant model to generate high-quality features as the supervision. These features are then used to train the LiDAR-Camera fusion model, where the fusion feature is optimized to simulate the generated high-quality features. Furthermore, we propose a simple yet effective deep fusion module, which contiguously gains superior performance compared with previous fusion methods with SupFusion strategy. In such a manner, our proposal shares the following advantages. Firstly, SupFusion introduces auxiliary feature-level supervision which could boost LiDAR-Camera detection performance without introducing extra inference costs. Secondly, the proposed deep fusion could continuously improve the detector's abilities. Our proposed SupFusion and deep fusion module is plug-and-play, we make extensive experiments to demonstrate its effectiveness. Specifically, we gain around 2% 3D mAP improvements on KITTI benchmark based on multiple LiDAR-Camera 3D detectors.
This paper presents two variations of architecture referred to as RANet and BIRANet. The proposed architecture aims to use radar signal data along with RGB camera images to form a robust detection network that works efficiently, even in variable lighting and weather conditions such as rain, dust, fog, and others. First, radar information is fused in the feature extractor network. Second, radar points are used to generate guided anchors. Third, a method is proposed to improve region proposal network targets. BIRANet yields 72.3/75.3% average AP/AR on the NuScenes dataset, which is better than the performance of our base network Faster-RCNN with Feature pyramid network(FFPN). RANet gives 69.6/71.9% average AP/AR on the same dataset, which is reasonably acceptable performance. Also, both BIRANet and RANet are evaluated to be robust towards the noise.
Self-supervised pretraining has been shown to yield powerful representations for transfer learning. These performance gains come at a large computational cost however, with state-of-the-art methods requiring an order of magnitude more computation than supervised pretraining. We tackle this computational bottleneck by introducing a new self-supervised objective, contrastive detection, which tasks representations with identifying object-level features across augmentations. This objective extracts a rich learning signal per image, leading to state-of-the-art transfer accuracy on a variety of downstream tasks, while requiring up to 10x less pretraining. In particular, our strongest ImageNet-pretrained model performs on par with SEER, one of the largest self-supervised systems to date, which uses 1000x more pretraining data. Finally, our objective seamlessly handles pretraining on more complex images such as those in COCO, closing the gap with supervised transfer learning from COCO to PASCAL.
In recent years, transformer-based detectors have demonstrated remarkable performance in 2D visual perception tasks. However, their performance in multi-view 3D object detection remains inferior to the state-of-the-art (SOTA) of convolutional neural network based detectors. In this work, we investigate this issue from the perspective of bird's-eye-view (BEV) feature generation. Specifically, we examine the BEV feature generation method employed by the transformer-based SOTA, BEVFormer, and identify its two limitations: (i) it only generates attention weights from BEV, which precludes the use of lidar points for supervision, and (ii) it aggregates camera view features to the BEV through deformable sampling, which only selects a small subset of features and fails to exploit all information. To overcome these limitations, we propose a novel BEV feature generation method, dual-view attention, which generates attention weights from both the BEV and camera view. This method encodes all camera features into the BEV feature. By combining dual-view attention with the BEVFormer architecture, we build a new detector named VoxelFormer. Extensive experiments are conducted on the nuScenes benchmark to verify the superiority of dual-view attention and VoxelForer. We observe that even only adopting 3 encoders and 1 historical frame during training, VoxelFormer still outperforms BEVFormer significantly. When trained in the same setting, VoxelFormer can surpass BEVFormer by 4.9% NDS point. Code is available at: https://github.com/Lizhuoling/VoxelFormer-public.git.
Salient Object Detection (SOD) has traditionally relied on feature refinement modules that utilize the features of an ImageNet pre-trained backbone. However, this approach limits the possibility of pre-training the entire network because of the distinct nature of SOD and image classification. Additionally, the architecture of these backbones originally built for Image classification is sub-optimal for a dense prediction task like SOD. To address these issues, we propose a novel encoder-decoder-style neural network called SODAWideNet++ that is designed explicitly for SOD. Inspired by the vision transformers ability to attain a global receptive field from the initial stages, we introduce the Attention Guided Long Range Feature Extraction (AGLRFE) module, which combines large dilated convolutions and self-attention. Specifically, we use attention features to guide long-range information extracted by multiple dilated convolutions, thus taking advantage of the inductive biases of a convolution operation and the input dependency brought by self-attention. In contrast to the current paradigm of ImageNet pre-training, we modify 118K annotated images from the COCO semantic segmentation dataset by binarizing the annotations to pre-train the proposed model end-to-end. Further, we supervise the background predictions along with the foreground to push our model to generate accurate saliency predictions. SODAWideNet++ performs competitively on five different datasets while only containing 35% of the trainable parameters compared to the state-of-the-art models. The code and pre-computed saliency maps are provided at https://github.com/VimsLab/SODAWideNetPlusPlus.
Multi-view camera-based 3D detection is a challenging problem in computer vision. Recent works leverage a pretrained LiDAR detection model to transfer knowledge to a camera-based student network. However, we argue that there is a major domain gap between the LiDAR BEV features and the camera-based BEV features, as they have different characteristics and are derived from different sources. In this paper, we propose Geometry Enhanced Masked Image Modeling (GeoMIM) to transfer the knowledge of the LiDAR model in a pretrain-finetune paradigm for improving the multi-view camera-based 3D detection. GeoMIM is a multi-camera vision transformer with Cross-View Attention (CVA) blocks that uses LiDAR BEV features encoded by the pretrained BEV model as learning targets. During pretraining, GeoMIM's decoder has a semantic branch completing dense perspective-view features and the other geometry branch reconstructing dense perspective-view depth maps. The depth branch is designed to be camera-aware by inputting the camera's parameters for better transfer capability. Extensive results demonstrate that GeoMIM outperforms existing methods on nuScenes benchmark, achieving state-of-the-art performance for camera-based 3D object detection and 3D segmentation. Code and pretrained models are available at https://github.com/Sense-X/GeoMIM.
Anomaly detection (AD) is a crucial machine learning task that aims to learn patterns from a set of normal training samples to identify abnormal samples in test data. Most existing AD studies assume that the training and test data are drawn from the same data distribution, but the test data can have large distribution shifts arising in many real-world applications due to different natural variations such as new lighting conditions, object poses, or background appearances, rendering existing AD methods ineffective in such cases. In this paper, we consider the problem of anomaly detection under distribution shift and establish performance benchmarks on three widely-used AD and out-of-distribution (OOD) generalization datasets. We demonstrate that simple adaptation of state-of-the-art OOD generalization methods to AD settings fails to work effectively due to the lack of labeled anomaly data. We further introduce a novel robust AD approach to diverse distribution shifts by minimizing the distribution gap between in-distribution and OOD normal samples in both the training and inference stages in an unsupervised way. Our extensive empirical results on the three datasets show that our approach substantially outperforms state-of-the-art AD methods and OOD generalization methods on data with various distribution shifts, while maintaining the detection accuracy on in-distribution data.
Due to the nature of enhancement--the absence of paired ground-truth information, high-level vision tasks have been recently employed to evaluate the performance of low-light image enhancement. A widely-used manner is to see how accurately an object detector trained on enhanced low-light images by different candidates can perform with respect to annotated semantic labels. In this paper, we first demonstrate that the mentioned approach is generally prone to overfitting, and thus diminishes its measurement reliability. In search of a proper evaluation metric, we propose LIME-Bench, the first online benchmark platform designed to collect human preferences for low-light enhancement, providing a valuable dataset for validating the correlation between human perception and automated evaluation metrics. We then customize LIME-Eval, a novel evaluation framework that utilizes detectors pre-trained on standard-lighting datasets without object annotations, to judge the quality of enhanced images. By adopting an energy-based strategy to assess the accuracy of output confidence maps, our LIME-Eval can simultaneously bypass biases associated with retraining detectors and circumvent the reliance on annotations for dim images. Comprehensive experiments are provided to reveal the effectiveness of our LIME-Eval. Our benchmark platform (https://huggingface.co/spaces/lime-j/eval) and code (https://github.com/lime-j/lime-eval) are available online.
In this work, we focus on semi-supervised learning for video action detection. Video action detection requires spatiotemporal localization in addition to classification, and a limited amount of labels makes the model prone to unreliable predictions. We present Stable Mean Teacher, a simple end-to-end teacher-based framework that benefits from improved and temporally consistent pseudo labels. It relies on a novel Error Recovery (EoR) module, which learns from students' mistakes on labeled samples and transfers this knowledge to the teacher to improve pseudo labels for unlabeled samples. Moreover, existing spatiotemporal losses do not take temporal coherency into account and are prone to temporal inconsistencies. To address this, we present Difference of Pixels (DoP), a simple and novel constraint focused on temporal consistency, leading to coherent temporal detections. We evaluate our approach on four different spatiotemporal detection benchmarks: UCF101-24, JHMDB21, AVA, and YouTube-VOS. Our approach outperforms the supervised baselines for action detection by an average margin of 23.5% on UCF101-24, 16% on JHMDB21, and 3.3% on AVA. Using merely 10% and 20% of data, it provides competitive performance compared to the supervised baseline trained on 100% annotations on UCF101-24 and JHMDB21, respectively. We further evaluate its effectiveness on AVA for scaling to large-scale datasets and YouTube-VOS for video object segmentation, demonstrating its generalization capability to other tasks in the video domain. Code and models are publicly available.
Fine-grained edited image detection of localized edits in images is crucial for assessing content authenticity, especially given that modern diffusion models and image editing methods can produce highly realistic manipulations. However, this domain faces three challenges: (1) Binary classifiers yield only a global real-or-fake label without providing localization; (2) Traditional computer vision methods often rely on costly pixel-level annotations; and (3) No large-scale, high-quality dataset exists for modern image-editing detection techniques. To address these gaps, we develop an automated data-generation pipeline to create FragFake, the first dedicated benchmark dataset for edited image detection, which includes high-quality images from diverse editing models and a wide variety of edited objects. Based on FragFake, we utilize Vision Language Models (VLMs) for the first time in the task of edited image classification and edited region localization. Experimental results show that fine-tuned VLMs achieve higher average Object Precision across all datasets, significantly outperforming pretrained models. We further conduct ablation and transferability analyses to evaluate the detectors across various configurations and editing scenarios. To the best of our knowledge, this work is the first to reformulate localized image edit detection as a vision-language understanding task, establishing a new paradigm for the field. We anticipate that this work will establish a solid foundation to facilitate and inspire subsequent research endeavors in the domain of multimodal content authenticity.
Video instance segmentation is a complex task in which we need to detect, segment, and track each object for any given video. Previous approaches only utilize single-frame features for the detection, segmentation, and tracking of objects and they suffer in the video scenario due to several distinct challenges such as motion blur and drastic appearance change. To eliminate ambiguities introduced by only using single-frame features, we propose a novel comprehensive feature aggregation approach (CompFeat) to refine features at both frame-level and object-level with temporal and spatial context information. The aggregation process is carefully designed with a new attention mechanism which significantly increases the discriminative power of the learned features. We further improve the tracking capability of our model through a siamese design by incorporating both feature similarities and spatial similarities. Experiments conducted on the YouTube-VIS dataset validate the effectiveness of proposed CompFeat. Our code will be available at https://github.com/SHI-Labs/CompFeat-for-Video-Instance-Segmentation.
The Lightweight Integrated Tracking-Feature Extraction (LITE) paradigm is introduced as a novel multi-object tracking (MOT) approach. It enhances ReID-based trackers by eliminating inference, pre-processing, post-processing, and ReID model training costs. LITE uses real-time appearance features without compromising speed. By integrating appearance feature extraction directly into the tracking pipeline using standard CNN-based detectors such as YOLOv8m, LITE demonstrates significant performance improvements. The simplest implementation of LITE on top of classic DeepSORT achieves a HOTA score of 43.03% at 28.3 FPS on the MOT17 benchmark, making it twice as fast as DeepSORT on MOT17 and four times faster on the more crowded MOT20 dataset, while maintaining similar accuracy. Additionally, a new evaluation framework for tracking-by-detection approaches reveals that conventional trackers like DeepSORT remain competitive with modern state-of-the-art trackers when evaluated under fair conditions. The code will be available post-publication at https://github.com/Jumabek/LITE.
RGB-D salient object detection (SOD) recently has attracted increasing research interest by benefiting conventional RGB SOD with extra depth information. However, existing RGB-D SOD models often fail to perform well in terms of both efficiency and accuracy, which hinders their potential applications on mobile devices and real-world problems. An underlying challenge is that the model accuracy usually degrades when the model is simplified to have few parameters. To tackle this dilemma and also inspired by the fact that depth quality is a key factor influencing the accuracy, we propose a novel depth quality-inspired feature manipulation (DQFM) process, which is efficient itself and can serve as a gating mechanism for filtering depth features to greatly boost the accuracy. DQFM resorts to the alignment of low-level RGB and depth features, as well as holistic attention of the depth stream to explicitly control and enhance cross-modal fusion. We embed DQFM to obtain an efficient light-weight model called DFM-Net, where we also design a tailored depth backbone and a two-stage decoder for further efficiency consideration. Extensive experimental results demonstrate that our DFM-Net achieves state-of-the-art accuracy when comparing to existing non-efficient models, and meanwhile runs at 140ms on CPU (2.2$\times$ faster than the prior fastest efficient model) with only $\sim$8.5Mb model size (14.9% of the prior lightest). Our code will be available at https://github.com/zwbx/DFM-Net.
We rethink the segment anything model (SAM) and propose a novel multiprompt network called COMPrompter for camouflaged object detection (COD). SAM has zero-shot generalization ability beyond other models and can provide an ideal framework for COD. Our network aims to enhance the single prompt strategy in SAM to a multiprompt strategy. To achieve this, we propose an edge gradient extraction module, which generates a mask containing gradient information regarding the boundaries of camouflaged objects. This gradient mask is then used as a novel boundary prompt, enhancing the segmentation process. Thereafter, we design a box-boundary mutual guidance module, which fosters more precise and comprehensive feature extraction via mutual guidance between a boundary prompt and a box prompt. This collaboration enhances the model's ability to accurately detect camouflaged objects. Moreover, we employ the discrete wavelet transform to extract high-frequency features from image embeddings. The high-frequency features serve as a supplementary component to the multiprompt system. Finally, our COMPrompter guides the network to achieve enhanced segmentation results, thereby advancing the development of SAM in terms of COD. Experimental results across COD benchmarks demonstrate that COMPrompter achieves a cutting-edge performance, surpassing the current leading model by an average positive metric of 2.2% in COD10K. In the specific application of COD, the experimental results in polyp segmentation show that our model is superior to top-tier methods as well. The code will be made available at https://github.com/guobaoxiao/COMPrompter.
The recently proposed Detection Transformer (DETR) model successfully applies Transformer to objects detection and achieves comparable performance with two-stage object detection frameworks, such as Faster-RCNN. However, DETR suffers from its slow convergence. Training DETR \cite{carion2020end} from scratch needs 500 epochs to achieve a high accuracy. To accelerate its convergence, we propose a simple yet effective scheme for improving the DETR framework, namely Spatially Modulated Co-Attention (SMCA) mechanism. The core idea of SMCA is to conduct regression-aware co-attention in DETR by constraining co-attention responses to be high near initially estimated bounding box locations. Our proposed SMCA increases DETR's convergence speed by replacing the original co-attention mechanism in the decoder while keeping other operations in DETR unchanged. Furthermore, by integrating multi-head and scale-selection attention designs into SMCA, our fully-fledged SMCA can achieve better performance compared to DETR with a dilated convolution-based backbone (45.6 mAP at 108 epochs vs. 43.3 mAP at 500 epochs). We perform extensive ablation studies on COCO dataset to validate the effectiveness of the proposed SMCA.