Anabranch network for camouflaged object segmentation

Camouflage objects hide information physically based on the feature matching of the texture or boundary line within the background. Texture matching and similarities between the camouflage objects and surrounding maps make differentiation difficult with generic and salient objects, thus making camouflage object detection (COD) more challenging. The existing techniques perform well. However, the challenging nature of camouflage objects demands more accuracy in detection and segmentation. To overcome this challenge, an optimized modular framework for COD tasks, named Optimize Global Refinement (OGR), is presented. This framework comprises a parallelism approach in feature extraction for the enhancement of learned parameters and globally refined feature maps for the abstraction of all intuitive feature sets at each extraction block’s outcome. Additionally, an optimized local best feature node-based rule is proposed to reduce the complexity of the proposed model. In light of the baseline experiments, OGR was applied and evaluated on a benchmark. The publicly available datasets were outperformed by achieving state-of-the-art structural similarity of 94%, 93%, and 96% for the Kvasir-SEG, COD10K, and Camouflaged Object (CAMO) datasets, respectively. The OGR is generalized and can be integrated into real-time applications for future development.

Camouflaged object detection (COD) is a new computer vision challenge for locating and identifying camouflaged objects in complex situations. Camouflaged objects are more similar to their surroundings than conventional objects, and their appearance in terms of size and shape is also considerably different, making accurate identification of the COD tasks difficult. As a result, we propose an enhanced identification network (EINet) to strengthen the COD task’s identification capabilities. First, the pyramid vision transformer is used as an encoder for extracting more robust multiscale features. Second, the multiple texture refinement modules are exploited to refine the multiscale features. Third, an improved neighbor and hop connection decoder is designed to produce a coarse estimation map for guiding the detailed identification of camouflaged objects backward. Finally, numerous new reverse criss-cross block attention modules that gradually recognize fine-grained features at various scales is designed to allow for the accurate recognition of camouflaged objects. Extensive experiments have been conducted on four benchmarked datasets of camouflaged objects. The results of the experiments reveal that our EINet is a powerful COD model that outperforms current state-of-the-art models.

Camouflaged objects, similar to the background, show indefinable boundaries and deceptive textures, which increases the difficulty of detection task and makes the model rely on features with more information. Herein, we design a texture label to facilitate our network for accurate camouflaged object segmentation. Motivated by the complementary relationship between texture labels and camouflaged object labels, we propose an interactive guidance framework named TINet, which focuses on finding the indefinable boundary and the texture difference by progressive interactive guidance. It maximizes the guidance effect of refined multi-level texture cues on segmentation. Specifically, texture perception decoder (TPD) makes a comprehensive analysis of texture information in multiple scales. Feature interaction guidance decoder (FGD) interactively refines multi-level features of camouflaged object detection and texture detection level by level. Holistic perception decoder (HPD) enhances FGD results by multi-level holistic perception. In addition, we propose a boundary weight map to help the loss function pay more attention to the object boundary. Sufficient experiments conducted on COD and SOD datasets demonstrate that the proposed method performs favorably against 23 state-of-the-art methods.

Existing camouflaged object detection (COD) methods rely heavily on large-scale datasets with pixel-wise annotations. However, due to the ambiguous boundary, annotating camouflage objects pixel-wisely is very time-consuming and labor-intensive, taking ~60mins to label one image. In this paper, we propose the first weakly-supervised COD method, using scribble annotations as supervision. To achieve this, we first relabel 4,040 images in existing camouflaged object datasets with scribbles, which takes ~10s to label one image. As scribble annotations only describe the primary structure of objects without details, for the network to learn to localize the boundaries of camouflaged objects, we propose a novel consistency loss composed of two parts: a cross-view loss to attain reliable consistency over different images, and an inside-view loss to maintain consistency inside a single prediction map. Besides, we observe that humans use semantic information to segment regions near the boundaries of camouflaged objects. Hence, we further propose a feature-guided loss, which includes visual features directly extracted from images and semantically significant features captured by the model. Finally, we propose a novel network for COD via scribble learning on structural information and semantic relations. Our network has two novel modules: the local-context contrasted (LCC) module, which mimics visual inhibition to enhance image contrast/sharpness and expand the scribbles into potential camouflaged regions, and the logical semantic relation (LSR) module, which analyzes the semantic relation to determine the regions representing the camouflaged object. Experimental results show that our model outperforms relevant SOTA methods on three COD benchmarks with an average improvement of 11.0% on MAE, 3.2% on S-measure, 2.5% on E-measure, and 4.4% on weighted F-measure.

Advances in herpetological research emanating from China.

Camouflaged Object Segmentation (COS) faces significant challenges due to the scarcity of annotated data, where meticulous pixel-level annotation is both labor-intensive and costly, primarily due to the intricate object-background boundaries. Addressing the core question, "Can COS be effectively achieved in a zero-shot manner without manual annotations for any camouflaged object?", we propose an affirmative solution. We examine the learned attention patterns for camouflaged objects and introduce a robust zero-shot COS framework. Our findings reveal that while transformer models for salient object segmentation (SOS) prioritize global features in their attention mechanisms, camouflaged object segmentation exhibits both global and local attention biases. Based on these findings, we design a framework that adapts with the inherent local pattern bias of COS while incorporating global attention patterns and a broad semantic feature space derived from SOS. This enables efficient zero-shot transfer for COS. Specifically, We incorporate a Masked Image Modeling (MIM) based image encoder optimized for Parameter-Efficient Fine-Tuning (PEFT), a Multimodal Large Language Model (M-LLM), and a Multi-scale Fine-grained Alignment (MFA) mechanism. The MIM encoder captures essential local features, while the PEFT module learns global and semantic representations from SOS datasets. To further enhance semantic granularity, we leverage the M-LLM to generate caption embeddings conditioned on visual cues, which are meticulously aligned with multi-scale visual features via MFA. This alignment enables precise interpretation of complex semantic contexts. Moreover, we introduce a learnable codebook to represent the M-LLM during inference, significantly reducing computational demands while maintaining performance. Our framework demonstrates its versatility and efficacy through rigorous experimentation, achieving state-of-the-art performance in zero-shot COS with $F_{\beta }^{w}$Fβw scores of 72.9% on CAMO and 71.7% on COD10K. By removing the M-LLM during inference, we achieve an inference speed comparable to that of traditional end-to-end models, reaching 18.1 FPS. Additionally, our method excels in polyp segmentation, and underwater scene segmentation, outperforming challenging baselines in both zero-shot and supervised settings, thereby implying its potentiality in various segmentation tasks.

Camouflaged object detection aims to identify objects that blend seamlessly with their background, posing a greater challenge compared to general object detection tasks. Due to its ability to recognize camouflaged objects, such detection models hold significant practical value across various fields. To accurately identify camouflaged targets in various complex environments, we designed a dual-guided camouflaged object detection network based on boundary and texture information(BTDGNet). The process consists of two main stages. The first stage is the localization stage, which leverages a convolutional neural network (CNN) to capture boundary and texture information of objects. These features are then fused to achieve coarse localization of the camouflaged objects. In the second stage, the recognition stage, we employ a Transformer to extract global information from the image, enhancing the differentiation between foreground and background. An interactive fusion module is designed to fully exploit and integrate both global and local features, producing precise prediction images. By leveraging boundary and texture information, the model's adaptability to different camouflaged objects is improved. The integration of local and global features enhances the model's detection accuracy from various perspectives, ultimately building a camouflaged object detection model suitable for a wide range of complex scenarios. The proposed method was extensively compared with other state-of-the-art methods across four public datasets, and the results demonstrated superior performance. Furthermore, benefiting from our dual-guidance strategy that leverages both texture and boundary information, our model demonstrates robust performance. We conducted tests on detection tasks across four different domains, and the results confirm that our model can accurately segment camouflaged objects in complex scenes.

IntroductionHow many species there are in the world remains a fundamental scientific question, serving as a critical reference for formulating and implementing effective biodiversity conservation strategies. The accelerating global biodiversity crisis has propelled scientific interest in understanding spatial-temporal patterns of new species discovery, particularly as these findings inform urgent conservation priorities. Digitalization of plant specimens provides important information on the discovery process of new species and clarifies the general situation of the core distributions in China over the last century. However, the new species discovery process of plants has not been the focus of much attention in Asia. The study analyzes digitized herbarium specimen data to investigate new species discoveries, specimen collection gaps, and collection preferences in China over the past century.MethodsFirst, we collected the herbarium type specimens data. Then we analyzed sampling biases of functional traits (life form, floral colors, fruit colors and types) in specimens collection and summarize distributional patterns in new species discovery. We answered the question of what plant species are more likely to be found and explored the distribution areas of new species discoveries over the last century.ResultsOur results reveal that there are more new plant species been found in herbaceous plants (90.8%) than woody (9.2%) plants, and new species with bright flowers and fruits, and the smaller height are easier to be discovered. There is also an observable inter-annual and seasonal variation in the discovery of new species, with peak collections occurring four year periods, and summer (June-August) witnessing the highest number of new species discoveries. The southern regions of China offer easier access to new species, and higher numbers of new species have been identified in the regions with the richness of plant diversity. The number of specimens collected by individuals (58%) surpasses those collected by teams (42%).DiscussionMore attention should also be paid to species with different plant functional traits (herbs, species with bright flowers and fruits) and the potential distribution of vacant regions. This study provides scientific reference and direction for the accelerating discovery of more new species in the future, and also contribute to the future conservation of biodiversity.

Open-vocabulary camouflaged object segmentation (OVCOS) seeks to segment and classify camouflaged objects in arbitrary categories, presenting unique challenges due to visual ambiguity and unseen categories. Recent approaches typically adopt a two-stage paradigm: they first segment objects, and then classify the segmented regions using vision language models (VLMs). However, such methods (i) suffer from a domain gap caused by the mismatch between VLMs' full-image training and cropped-region inferencing, and (ii) depend on generic segmentation models optimized for well-delineated objects which are less effective for camouflaged objects. Without explicit guidance, generic segmentation models often overlook subtle boundaries, leading to imprecise segmentation. In this paper, we introduce a novel VLM-guided cascaded framework to address these issues in OVCOS. For segmentation, we leverage the segment anything model (SAM), guided by the VLM. Our framework uses VLM-derived features as explicit prompts to SAM, effectively directing attention to camouflaged regions and significantly improving localization accuracy. For classification, we avoid the domain gap introduced by hard cropping. Instead, we treat the segmentation output as a soft spatial prior using the alpha channel. This retains the full image context while providing precise spatial guidance, leading to more accurate and context-aware classification of camouflaged objects. The same VLM is shared between segmentation and classification to ensure efficiency and semantic consistency. Extensive experiments on both OVCOS and conventional camouflaged object segmentation benchmarks demonstrate the clear superiority of our method, highlighting the effectiveness of leveraging rich VLM semantics for both segmentation and classification of camouflaged objects. Our code and models are open-sourced at https://github.com/intcomp/camouflaged-vlm.

In recent years, protecting important objects by simulating animal camouflage has been widely employed in many fields. Therefore, camouflaged object detection (COD) technology has emerged. COD is more difficult to achieve than traditional object detection techniques due to the high degree of fusion of objects camouflaged with the background. In this paper, we strive to more accurately and efficiently identify camouflaged objects. Inspired by the use of magnifiers to search for hidden objects in pictures, we propose a COD network that simulates the observation effect of a magnifier called the MAGnifier Network (MAGNet). Specifically, our MAGNet contains two parallel modules: the ergodic magnification module (EMM) and the attention focus module (AFM). The EMM is designed to mimic the process of a magnifier enlarging an image, and AFM is used to simulate the observation process in which human attention is highly focused on a particular region. The two sets of output camouflaged object maps were merged to simulate the observation of an object by a magnifier. In addition, a weighted key point area perception loss function, which is more applicable to COD, was designed based on two modules to give greater attention to the camouflaged object. Extensive experiments demonstrate that compared with 19 cutting-edge detection models, MAGNet can achieve the best comprehensive effect on eight evaluation metrics in the public COD dataset. Additionally, compared to other COD methods, MAGNet has lower computational complexity and faster segmentation. We also validated the model's generalization ability on a military camouflaged object dataset constructed in-house. Finally, we experimentally explored some extended applications of COD.

Novel network designs for deep learning show promise in overcoming the inherent similarity between hidden objects and the background environment, which hinders automatic detection and segmentation of camouflaged objects and improves detection accuracy. Nevertheless, the current network topologies for differentiating between camouflaged objects and their backgrounds need to consider the limitation of detection speed, leading to increased computational complexity and an inability to fulfil the demands of quick detection. This research proposes a new deep learning-based approach, Enhanced Multi-Head Self-Attention Graph Neural Networks (EMSAGNN), to detect and classify camouflage objects. In the pre-processing stage, noise reduction will be performed using a median filter, colour normalization, and contrast enhancement using the Adaptively Clipped Contrast Limited Adaptive Histogram Equalization (ACCLAHE) method. Then, extract the relevant features such as colour, shape, and texture from a pre-processed image using the ResNeXt50 technique. After that, detect the camouflaged object from the pre-processed input image with extracted features using the YOLOv8 detection method. Then, the Improved African Vulture Segmentation Algorithm (IAVSA) will be utilized to segment the identified camouflage objects. Finally, the novel deep learning technique is employed to classify the camouflage objects into their respective classes. According to extensive tests conducted on four challenging benchmark datasets: CHAMELEON, CAMO, COD10K, and NC4K, the proposed approach performs better than many cutting-edge techniques in terms of accuracy and model parameters.

We consider the problem of referring camouflaged object detection (Ref-COD), a new task that aims to segment specified camouflaged objects based on a small set of referring images with salient target objects. We first assemble a large-scale dataset, called R2C7K, which consists of 7 K images covering 64 object categories in real-world scenarios. Then, we develop a simple but strong dual-branch framework, dubbed R2CNet, with a reference branch embedding the common representations of target objects from referring images and a segmentation branch identifying and segmenting camouflaged objects under the guidance of the common representations. In particular, we design a Referring Mask Generation module to generate pixel-level prior mask and a Referring Feature Enrichment module to enhance the capability of identifying specified camouflaged objects. Extensive experiments show the superiority of our Ref-COD methods over their COD counterparts in segmenting specified camouflaged objects and identifying the main body of target objects.

Weakly Supervised Camouflaged Object Detection (WS-COD) aims to locate camouflaged objects with only sparse supervision, thereby substantially reducing the reliance on costly pixel-level annotations. This task poses two major challenges: limited supervision arising from sparse annotations (e.g., scribbles), and weak discriminability due to the inherent high visual similarity between camouflaged objects and their surroundings. To tackle these challenges, this paper proposes a novel multi-scale feature correlation transformer guided by the Segment Anything Model (SAM) for scribble-based WSCOD. Specifically, we introduce a cross-scale correlation module built upon Transformers, which exploits enriched cross-attention mechanisms to capture long-range global correlations and multi-scale discriminative cues, enabling accurate segmentation of camouflaged objects. In addition, we develop a SAM-based pseudo-label generation module that leverages sparse annotations as prompts to produce high-quality object masks, thereby enhancing supervision. Extensive experiments on three challenging datasets demonstrate that our proposed method consistently and significantly surpasses existing state-of-the-art approaches for scribble-based WSCOD. The code will be available at: http://github.com/ farewellIamLoser/FCT-SAM-WSCOD.

Camouflaged Object Detection (COD) is a challenging computer vision task aimed at accurately identifying and segmenting objects seamlessly blended into their backgrounds. This task has broad applications across medical image segmentation, defect detection, agricultural image detection, security monitoring, and scientific research. Traditional COD methods often struggle with precise segmentation due to the high similarity between camouflaged objects and their surroundings. In this study, we introduce a Boundary-Guided Differential Attention Network (BDA-Net) to address these challenges. BDA-Net first extracts boundary features by fusing multi-scale image features and applying channel attention. Subsequently, it employs a differential attention mechanism, guided by these boundary features, to highlight camouflaged objects and suppress background information. The weighted features are then progressively fused to generate accurate camouflage object masks. Experimental results on the COD10K, NC4K, and CAMO datasets demonstrate that BDA-Net outperforms most state-of-the-art COD methods, achieving higher accuracy. Here we show that our approach improves detection accuracy by up to 3.6% on key metrics, offering a robust solution for precise camouflaged object segmentation.

Camouflaged Object Detection (COD) is a challenging task in computer vision due to the high visual similarity between camouflaged objects and their surrounding environments. Traditional methods relying on the late-stage fusion of high-level semantic features and low-level visual features have reached a performance plateau, limiting their ability to accurately segment object boundaries or enhance object localization. This paper proposes the Cross-layer Semantic Guidance Network (CSGNet), a novel framework designed to progressively integrate semantic and visual features across multiple stages, addressing these limitations. CSGNet introduces two innovative modules: the Cross-Layer Interaction Module (CLIM) and the Semantic Refinement Module (SRM). CLIM facilitates continuous cross-layer semantic interaction, refining high-level semantic information to provide consistent and effective guidance for detecting camouflaged objects. Meanwhile, SRM leverages this refined semantic guidance to enhance low-level visual features, employing feature-level attention mechanisms to suppress background noise and highlight critical object details. This progressive integration strategy ensures precise object localization and accurate boundary segmentation across challenging scenarios. Extensive experiments on three widely used COD benchmark datasets—CAMO, COD10K, and NC4K—demonstrate the effectiveness of CSGNet, achieving state-of-the-art performance with a mean error (M) of 0.042 on CAMO, 0.020 on COD10K, and 0.029 on NC4K.