Image Features Research Articles

An insight to PDAC tumor heterogeneity across pancreatic subregions using computed tomography images

Pancreatic Ductal Adenocarcinoma (PDAC) is an exceptionally deadly form of pancreatic cancer with an extremely low survival rate. From diagnosis to treatment, PDAC is highly challenging to manage. Studies have demonstrated that PDAC tumors in distinct regions of the pancreas exhibit unique characteristics, influencing symptoms, treatment responses, and survival rates. Gaining insight into the heterogeneity of PDAC tumors based on their location in the pancreas can significantly enhance overall management of PDAC. Previous studies have explored PDAC tumor heterogeneity across pancreatic subregions based on their genetic and molecular profiles through biopsy-based histologic assessment. However, biopsy examinations are highly invasive and impractical for large populations. Abdominal imaging, such as Computed Tomography (CT) offers a completely non-invasive means to evaluate PDAC tumor heterogeneity across pancreatic subregions and an opportunity to correlate image feature of tumors with treatment outcome and monitoring. In this study, we explored the inter-tumor heterogeneity in PDAC tumors across three primary pancreatic subregions: the head, body, and tail. Utilizing contrast-enhanced abdominal CT scans and a thorough radiomic analysis of PDAC tumors, several morphological and textural tumor features were identified to be notably different between tumors in the head and those in the body and tail regions. To validate the significance of the identified features, a machine learning ML model was trained to automatically classify PDAC tumors into their respective regions i.e. head or body/tail subregion using their CT features. The study involved 200 CT abdominal scans, with 100 used for radiomic analysis and model training, and the remaining 100 for model testing. The ML model achieved an average classification accuracy, sensitivity, and specificity of 87%, 86%, and 88% on the testing scans respectively. Evaluating the heterogeneity of PDAC tumors across pancreatic subregions provides valuable insights into tumor composition and has the potential to enhance diagnosis and personalize treatment based on tumor characteristics and location.

NIMG-87. MULTIREGIONAL, MRI-LOCALIZED SAMPLING OF HIGH GRADE GLIOMAS REVEAL SEX-DISTINCT PATTERNS OF 5-ALA FLUORESCENCE AND IMAGING FEATURES

Abstract High grade gliomas (HGGs) are aggressive brain tumors that are difficult to remove completely, leading to poor survival rates. Contrast enhancement on T1Gd magnetic resonance imaging (MRI) has traditionally been used to guide resection, however it often underestimates true tumor burden, adding a layer of uncertainty for surgeons during pre-operative planning. Fluorescence guided surgery using 5-aminolevulinic acid (5-ALA) has emerged as an intraoperative solution and understanding it in conjunction with MRI features can address the limitations of routine T1Gd imaging during surgical planning. Here we investigate sex-specific patterns in HGGs using 5-ALA fluorescence and imaging features. We analyzed a patient dataset of 202 multiregional MRI-localized biopsies from 42 HGG patients (20 male and 22 female). Data included patient demographics, 5-ALA status (positive or negative), MRI annotations (CE = contrast enhancement, NE = non-enhancement, BAT = brain around tumor non-enhancement), and survival outcomes. Statistical analyses included Wilcoxon rank-sum tests, Kruskal-Wallis tests, and chi-squared tests to assess differences in imaging annotations and 5-ALA positive vs negative status among sex cohorts. Among all samples, CE and BAT were associated with higher 5-ALA positivity, while NE was associated with higher 5-ALA negativity (p = 0.02119). Female patients were less likely to have detectable 5-ALA in their multiregional biopsy samples than males (p < 0.0001). CE at the biopsy location was more likely to be indicative of 5-ALA positivity in males than females (p = 0.001927); whereas NE was associated with higher 5-ALA negativity in females than males (p = 0.03426). BAT was associated with higher 5-ALA positivity in males than females (p = 0.03895). Our study reveals sex-specific patterns in 5-ALA fluorescence and image features among HGG patients. These unique features may serve as potential pre-surgical markers for tumor regions that are more likely to benefit from 5-ALA-guided resection.

Postoperative Karnofsky performance status prediction in patients with IDH wild-type glioblastoma: A multimodal approach integrating clinical and deep imaging features.

Glioblastoma is a highly aggressive brain tumor with limited survival that poses challenges in predicting patient outcomes. The Karnofsky Performance Status (KPS) score is a valuable tool for assessing patient functionality and contributes to the stratification of patients with poor prognoses. This study aimed to develop a 6-month postoperative KPS prediction model by combining clinical data with deep learning-based image features from pre- and postoperative MRI scans, offering enhanced personalized care for glioblastoma patients. Using 1,476 MRI datasets from the Brain Tumor Segmentation Challenge 2020 public database, we pretrained two variational autoencoders (VAEs). Imaging features from the latent spaces of the VAEs were used for KPS prediction. Neural network-based KPS prediction models were developed to predict scores below 70 at 6 months postoperatively. In this retrospective single-center analysis, we incorporated clinical parameters and pre- and postoperative MRI images from 150 newly diagnosed IDH wild-type glioblastoma, divided into training (100 patients) and test (50 patients) sets. In training set, the performance of these models was evaluated using the area under the curve (AUC), calculated through fivefold cross-validation repeated 10 times. The final evaluation of the developed models assessed in the test set. Among the 150 patients, 61 had 6-month postoperative KPS scores below 70 and 89 scored 70 or higher. We developed three models: a clinical-based model, an MRI-based model, and a multimodal model that incorporated both clinical parameters and MRI features. In the training set, the mean AUC was 0.785±0.051 for the multimodal model, which was significantly higher than the AUCs of the clinical-based model (0.716±0.059, P = 0.038) using only clinical parameters and the MRI-based model (0.651±0.028, P<0.001) using only MRI features. In the test set, the multimodal model achieved an AUC of 0.810, outperforming the clinical-based (0.670) and MRI-based (0.650) models. The integration of MRI features extracted from VAEs with clinical parameters in the multimodal model substantially enhanced KPS prediction performance. This approach has the potential to improve prognostic prediction, paving the way for more personalized and effective treatments for patients with glioblastoma.

Combined query embroidery image retrieval based on enhanced CNN and blend transformer.

Embroidery images carry rich historical information and are an important form of embroidery art. In the field of combination query image retrieval, how to efficiently retrieve the embroidery image information required by users has become a current research challenge. In recent years, convolutional neural networks (CNNs) have achieved significant success in image feature extraction, but they tend to focus on local information, making it easy to ignore global context information when processing such textured embroidery images. Therefore, we propose a combination query retrieval method for embroidery images. First, we propose Blend-Transformer, which introduces Group External Attention (GEA). GEA can integrate feature information from three different dimensions, effectively capturing the local and global context information of embroidery images. Second, we propose Enhanced CNN, which introduces Shuffle Attention (SA), regrouping the reference image features extracted by CNN and reaggregating them by channel to enhance the richness of embroidery image feature information. Through experiments on the TCE-S and ICR2020 standard datasets, we verify the excellent performance of the proposed algorithm in embroidery image retrieval. Our method fills the gap in embroidery image retrieval research and provides a new perspective for the protection of embroidery art.

Semantic Segmentation of Corn Leaf Blotch Disease Images Based on U-Net Integrated with RFB Structure and Dual Attention Mechanism

Northern corn leaf blight (NCLB) is caused by a fungus and can be susceptible to the disease throughout the growing period of corn, posing a significant impact on corn yield. Aiming at the problems of under-segmentation, over-segmentation, and low segmentation accuracy in the traditional segmentation model of northern corn leaf blight, this study proposes a segmentation method based on an improved U-Net network model. By introducing a convolutional layer and maximum pooling layer to a VGG19 network, the channel attention module and spatial attention module (CBAM) are fused, and the squeeze excitation (SE) attention mechanism is combined. This enhances image feature decoding, integrates feature maps of each layer, strengthens the feature extraction process, expands the sensory fields and aggregates context information, and reduces the loss of location and dense semantic information caused by the pooling operation. Findings from the study show that the proposed NCLB-Net has significantly improved the MIoU and PA indexes, reaching 92.43% and 94.71%, respectively. Compared with the traditional methods, U-Net, SETR, DAnet, OCnet, PSPNet, etc., the MIoU is improved by 20.81%, 16.10%, 9.79%, 5.27%, and 11.06%, and the PA is improved by 11.49%, 8.18%, 9.54%, 13.11%, and 6.26%, respectively.

Art Pattern Design Based on Visual Semantic Segmentation and Evolutionary Algorithm Optimization

This paper aims to address the issue of traditional art patterns not fully leveraging contextual information in image semantic segmentation. It proposes a multi-level feature aggregation method for image semantic segmentation based on visual Transformer and coordinate attention adjustment (CA Adjustment). Initially, the input image is segmented into slices for linear projection, incorporating learnable positional embeddings to derive an encoded input sequence. Subsequently, employing a visual Transformer-based encoder, the image is transformed into patches to capture global context across the network. CA Adjustment is then introduced to adaptively merge fusion features and visual backbone features using weighted adjustments. Lastly, a multi-objective grasshopper optimization algorithm is designed based on a clustering evolution mechanism to further enhance algorithm performance. Extensive experimental results demonstrate that our proposed algorithm effectively models global contextual information for image feature extraction. Compared to existing advanced algorithms, our method achieves high segmentation accuracy in semantic segmentation tasks, consistently exceeding 95%. Ablation experiments further validate the efficacy of our approach, underscoring its broad applicability in high-precision image semantic segmentation.

Neuroimage-Based Stroke Identification: A Machine Learning Approach

Stroke diagnosis is a time-critical process that requires rapid and accurate identification to ensure timely treatment. This study proposes a machine learning-based diagnostic model for stroke identification using neuro images. Early identification and timely intervention are critical to improving outcomes for stroke patients, but current diagnostic techniques, such as CT and MRI scans, often require time-consuming expert analysis. These delays can limit the effectiveness of treatment, particularly in acute cases where every minute counts. The problem lies in the need for faster, more reliable diagnostic tools that can analyze neuroimaging data with high accuracy and minimal human intervention. Machine learning, specifically deep learning, offers a promising solution to address this gap by automating the process of stroke detection. We employed a comprehensive approach, utilizing Inceptionv3, MobileNet, Convolutional Neural Network (CNN) algorithms to analyze neuroimages and predict stroke occurrence. This research proposes a machine learning-based diagnostic model for stroke identification using neuroimages, leveraging the power of Convolutional Neural Networks (CNN), with Inception V3 and MobileNet architectures. Inception V3, known for its ability to capture intricate image features through deep convolutional layers, and MobileNet, optimized for efficiency and speed, were employed to process large datasets of brain scans. The model was trained on these neuroimaging datasets to distinguish between healthy brain tissues and those affected by stroke. The combination of these two architectures allows for both detailed analysis and fast processing, making the model adaptable to clinical settings. The results showed that the model achieved a high accuracy rate in stroke identification, demonstrating its potential to assist healthcare professionals in diagnosing stroke faster and more accurately. By integrating this machine learning model into existing diagnostic workflows, it could significantly reduce the time to diagnosis, enabling earlier treatment and ultimately improving patient outcomes. Our model has the potential to enhance patient outcomes and reduce the economic burden of stroke. By leveraging the power of these advanced machine learning techniques, the model aims to enhance the efficiency and accuracy of stroke diagnosis compared to traditional methods

Machine Learning Based Framework for Lung Cancer Detection and Image Feature Extraction Using VGG16 with PCA on CT-Scan Images

Machine Learning Based Framework for Lung Cancer Detection and Image Feature Extraction Using VGG16 with PCA on CT-Scan Images

EPCO-29. PATHOMICS ANALYSIS IDENTIFIES POTENTIAL SUBTYPES OF DIFFUSE GLIOMAS IN H&amp;E-STAINED WHOLE-SLIDE IMAGES

Abstract IDH status is the first criteria of diagnostic definition for diffuse gliomas in the 2021 WHO classification. While sufficient number of molecular tools are available to identify IDH1 and/or IDH2 mutations, they impose a time and cost burden on patients. We previously trained deep learning models visually differentiate IDH wild-type (IDHwt) and IDH mutant (IDHmut) gliomas in a small set of H&amp;E-stained whole image slides (WSIs) from TCGA using an Uncertainty-Aware Convolutional Neural Network (CNN) at 10x magnification. In this follow-up study, we aimed to improve our models by expanding our annotations of a more representative subset of the TCGA public data, increasing the sampling magnification to 20x, redesigning the preprocessing filters, adjusting the training pipeline to reduce overfitting, and assessing the confidence in model prediction via uncertainty metrics. METHODS By modifying the pre-processing methods from our pilot study, we increased the number of expert-annotated WSIs. Using a 2:1:2 train/validation/test split, we fine-tuned a ResNet18 deep learning model with a pre-trained feature extractor and Monte Carlo dropouts. To identify image features critical for separating by IDH status, we extracted the learned embeddings and performed post-hoc analyses (PCA, t-SNE, and UMAP). RESULTS The new set of annotations resulted in ~15-fold increase in the number of WSIs (i.e.,&amp;gt;2.5 million tiles), without any data augmentation. Our prototype model demonstrated clear separation of cases according to IDH status (sensitivity = 0.63, specificity = 0.91, weighted F1 = 0.81). CONCLUSION We successfully improved the prediction of IDH status in H&amp;E-stained WSIs and identified potential subtypes of IDHwt and IDHmut diffuse gliomas after significantly increasing the dataset size with appropriate expert annotations. We are currently addressing the data imbalance by annotating additional IDHmut glioma WSIs. Furthermore, we will integrate other biomarkers (1p19q, EGFR etc.) to increase the robustness of histo-pathomics as a surrogate for molecular tests.

Pulmonary CT Registration Network Based on Deformable Cross Attention.

Current Transformer structure utilizes the self-attention mechanism to model global contextual relevance within image, which makes an impact on medical image registration. However, the use of Transformer in handling large deformation lung CT registration is relatively straightforwardly. These models only focus on single image feature representation neglecting to employ attention mechanism to capture the across image correspondence. This hinders further improvement in registration performance. To address the above limitations, we propose a novel registration method in a cascaded manner, Cascaded Swin Deformable Cross Attention Transformer based U-shape structure (SD-CATU), to address the challenge of large deformations in lung CT registration. In SD-CATU, we introduce a Cross Attention-based Transformer (CAT) block that incorporates the Shifted Regions Multihead Cross-attention (SR-MCA) mechanism to flexibly exchange feature information and thus reduce the computational complexity. Besides, a consistency constraint in the loss function is used to ensure the preservation of topology and inverse consistency of the transformations. Experiments with public lung datasets demonstrate that the Cascaded SD-CATU outperforms current state-of-the-art registration methods (Dice Similarity Coefficient of 93.19% and Target registration error of 0.98mm). The results further highlight the potential for obtaining excellent registration accuracy while assuring desirable smoothness and consistency in the deformed images.

Smart Caption: Intelligent Image Description Using Transformer Decoder

The need for automated image description systems has grown significantly with the rise of multimedia content across diverse domains such as social media, digital libraries, and accessibility technologies. Smart Caption presents a novel approach for generating intelligent and context-aware image descriptions by utilizing a combination of Vision Transformers (ViTs) and Natural Language Processing (NLP) Transformer decoders. Unlike traditional convolution-based methods, Vision Transformers treat images as sequences of patches, enabling them to capture global image features more effectively. These extracted visual features are then processed by a Transformer-based decoder to produce coherent and contextually appropriate captions, bridging the gap between image recognition and natural language generation. Our approach focuses on leveraging the attention mechanisms inherent in both Vision and NLP Transformers to enhance the quality of image descriptions. The ViT architecture is designed to focus on relevant regions within an image, while the NLP Transformer decoder is adept at generating fluent and detailed descriptions by attending to the most significant visual features. This two-stage process improves the precision of object detection, relationship understanding, and scene context in generated captions. Experimental results demonstrate that Smart Caption out performs traditional methods in terms of accuracy, contextual relevance, and fluency, marking a significant step forward in the field of automated image captioning. Through this research, we aim to provide a scalable, efficient, and intelligent solution for image description, with potential applications in areas such as accessibility tools for visually impaired users, digital content indexing, and automated content generation. Our findings highlight the strengths of transformer-based models in bridging vision and language tasks, offering promising directions for future research in multimodal AI..

A multi-verse Optimizer Approach Based on a Salp Swarm for Image Feature Selection to Observe Changes in Urban Lands

Many changes have occurred in most urban areas, and it is usually governed by many geographical and social factors, the result will be either the expansion or contraction of these cities. In this paper, an optimizer method called salp swarm optimization (SSO) is used to build a driving force model for urban lands according to the changes in the driving force mechanism. The principle of (SSO) is extracting the driving force classification rules corresponding to different types of land use change samples by imitating the behavior of salp. The classification rules are constructed in the form of "IF ... THEN", and three different fitness functions are selected for simulation verification. The data set used in this search represents images from Global Positioning System (GPS) satellites and remote sensing data. The experimental results show that the overall accuracy evaluation of the salp swarm optimization model is superior to other algorithms, indicating that the SSA is feasible for land use change modeling.

Mf-net: multi-feature fusion network based on two-stream extraction and multi-scale enhancement for face forgery detection

Due to the increasing sophistication of face forgery techniques, the images generated are becoming more and more realistic and difficult for human eyes to distinguish. These face forgery techniques can cause problems such as fraud and social engineering attacks in facial recognition and identity verification areas. Therefore, researchers have worked on face forgery detection studies and have made significant progress. Current face forgery detection algorithms achieve high detection accuracy within-dataset. However, it is difficult to achieve satisfactory generalization performance in cross-dataset scenarios. In order to improve the cross-dataset detection performance of the model, this paper proposes a multi-feature fusion network based on two-stream extraction and multi-scale enhancement. First, we design a two-stream feature extraction module to obtain richer feature information. Secondly, the multi-scale feature enhancement module is proposed to focus the model more on information related to the current sub-region from different scales. Finally, the forgery detection module calculates the overlap between the features of the input image and real images during the training phase to determine the forgery regions. The method encourages the model to mine forgery features and learns generic and robust features not limited to a particular feature. Thus, the model achieves high detection accuracy and performance. We achieve the AUC of 99.70% and 90.71% on FaceForensics++ and WildDeepfake datasets. The generalization experiments on Celeb-DF-v2 and WildDeepfake datasets achieve the AUC of 80.16% and 65.15%. Comparison experiments with multiple methods on other benchmark datasets confirm the superior generalization performance of our proposed method while ensuring model detection accuracy. Our code can be found at: https://github.com/1241128239/MFNet.

Enhancing Skin Lesion Classification Performance with the ABC Ensemble Model

Skin cancer is one of the most easily developed cancers and is continuously seeing an increased incidence rate. In this study, we propose a novel ABC ensemble model for skin lesion classification by leveraging the ABCD rule, which is commonly used in dermatology to evaluate lesion features such as asymmetry, border, color, and diameter. Our model consists of five distinct blocks, two of which focus on learning general image characteristics, while the remaining three focus on specialized features related to the ABCD rule. The final classification results are achieved through a weighted soft voting approach, where the generalization blocks are assigned higher weights to optimize performance. Through 15 experiments using various model configurations, we show that the weighted ABC ensemble model outperforms the baseline models, achieving the best performance with an accuracy of 0.9326 and an F1-score of 0.9302. Additionally, Grad-CAM analysis is employed to assess how each block in the ensemble focuses on distinct lesion features, further enhancing the interpretability and reliability of the model. Our findings demonstrate that integrating general image features with specific lesion characteristics improves classification performance, and that adjusting the soft voting weights yields optimal results. This novel model offers a reliable tool for early skin lesion diagnosis.

Photometric Redshift Estimation of Quasars by a Cross-modal Contrast Learning Method

Estimating photometric redshifts (photo-z) of quasars is crucial for measuring cosmic distances and monitoring cosmic evolution. While numerous point estimation methods have successfully determined photo-z, they often struggle with the inherently ill-posed nature of the problem and frequently overlook significant morphological features in the probability density functions (pdfs) of photo-z, such as calibration and sharpness. To address these challenges, we introduce a cross-modal contrastive learning probabilistic model that employs adversarial training, contrastive loss functions, and a mixture density network to estimate the pdf of photo-z. This method facilitates the conversion between multiband photometric data attributes, such as magnitude and color, and photometric image features, while extracting features invariant across modalities. We utilize the continuous ranked probability score (CRPS) and the probability integral transform (PIT) as metrics to assess the quality of the pdf. Our approach demonstrates robust performance across various survey bands, image qualities, and redshift distributions. Specifically, in a comprehensive data set from the Sloan Digital Sky Survey and the Wide-field Infrared Survey Explorer (WISE) survey, our probabilistic model achieved a CRPS of 0.1187. Additionally, in a combined data set from SkyMapper and WISE, it reached a CRPS of 0.0035. Our probabilistic model also produced well-calibrated PIT histograms for both data sets, indicating nearly uniform distributions. We further tested our approach in classification tasks within the SkyMapper data set. Despite the absence of u, v, and g bands, it effectively distinguished between quasars, galaxies, and stars with an accuracy of 98.96%. This versatile method can be extended to other scenarios, such as analyzing extended sources like galaxies, across different surveys and varying redshift distributions.

Attention-enhanced multimodal feature fusion network for clothes-changing person re-identification

Clothes-Changing Person Re-Identification is a challenging problem in computer vision, primarily due to the appearance variations caused by clothing changes across different camera views. This poses significant challenges to traditional person re-identification techniques that rely on clothing features. These challenges include the inconsistency of clothing and the difficulty in learning reliable clothing-irrelevant local features. To address this issue, we propose a novel network architecture called the Attention-Enhanced Multimodal Feature Fusion Network (AE-Net). AE-Net effectively mitigates the impact of clothing changes on recognition accuracy by integrating RGB global features, grayscale image features, and clothing-irrelevant features obtained through semantic segmentation. Specifically, global features capture the overall appearance of the person; grayscale image features help eliminate the interference of color in recognition; and clothing-irrelevant features derived from semantic segmentation enforce the model to learn features independent of the person’s clothing. Additionally, we introduce a multi-scale fusion attention mechanism that further enhances the model’s ability to capture both detailed and global structures, thereby improving recognition accuracy and robustness. Extensive experimental results demonstrate that AE-Net outperforms several state-of-the-art methods on the PRCC and LTCC datasets, particularly in scenarios with significant clothing changes. On the PRCC and LTCC datasets, AE-Net achieves Top-1 accuracy rates of 60.4% and 42.9%, respectively.

Enhancing Movie Recommendation Systems Through CNN-Based Feature Extraction and Optimized Collaborative Filtering

Abstract. This paper presents the design and implementation of a movie recommendation model that combines collaborative filtering with Convolutional Neural Networks (CNNs) to tackle the challenge of information overload in extensive movie databases. The primary objective is to enhance the user experience by developing an efficient and personalized recommendation system. The proposed model integrates CNNs to extract detailed image features from movie posters, enriching the movie representations used in the recommendation process. This feature extraction is coupled with an optimized collaborative filtering algorithm to deliver more accurate and tailored movie suggestions. The study is based on a dataset that includes movie information, user ratings, and user behavior data. Experimental results demonstrate that the proposed model significantly improves the accuracy and personalization of movie recommendations, thereby offering users a superior selection experience. The practical implications of this research are substantial, as it has the potential to boost user satisfaction and engagement on movie recommendation platforms. Future work will focus on further enhancing the model by incorporating long-term user behavior patterns and integrating multi-modal data sources for even more precise recommendations.

Enhancing Black-and-White Image Colorization Using Cycle GAN: A Study on Unpaired Image Translation

Abstract. This paper explores the application of Cycle-Consistent Generative Adversarial Network (Cycle GAN) for the colorization of black-and-white images, addressing the challenges in image to the image translation and it was not requiring paired datasets. Cycle GAN, a generative adversarial network, uses cycle consistency loss to map grayscale images to color, effectively preserving structural integrity. The model is trained on unpaired datasets containing black-and-white and color images, enabling it to generate colorized versions from grayscale inputs. Performance evaluation is conducted using Two significant evaluation metrics: Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index (SSIM). Results demonstrate that Cycle GAN effectively preserves image structure and detail, achieving high PSNR and SSIM values, particularly in scenes with distinct color boundaries. However, challenges arise in complex and low-resolution scenes where color accuracy and detail reconstruction may degrade. Additionally, this paper assesses the impact of different loss functions, learning rate schedules, and training strategies on the models output quality. Future work will focus on addressing these limitations by incorporating attention mechanisms and other advanced techniques to improve Cycle GAN's ability to handle intricate image features and complex scenes.

Impact of Sound and Image Features in ASMR on Emotional and Physiological Responses

Impact of Sound and Image Features in ASMR on Emotional and Physiological Responses

Visual software defect prediction method based on improved recurrent criss-cross residual network

Purpose This study aims to solve the problems of large training sample size, low data sample quality, low efficiency of the currently used classical model, high computational complexity of the existing concern mechanism, and high graphics processing unit (GPU) occupancy in the current visualization software defect prediction, proposing a method for software defect prediction termed recurrent criss-cross attention for weighted activation functions of recurrent SE-ResNet (RCCA-WRSR). First, following code visualization, the activation functions of the SE-ResNet model are replaced with a weighted combination of Relu and Elu to enhance model convergence. Additionally, an SE module is added before it to filter feature information, eliminating low-weight features to generate an improved residual network model, WRSR. To focus more on contextual information and establish connections between a pixel and those not in the same cross-path, the visualized red as integer, green as integer, blue as integer images are inputted into a model incorporating a fused RCCA module for defect prediction. Design/methodology/approach Software defect prediction based on code visualization is a new software defect prediction technology, which mainly realizes the defect prediction of code by visualizing code as image, and then applying attention mechanism to extract the features of image. However, the challenges of current visualization software defect prediction mainly include the large training sample size and low sample quality of the data, and the classical models used today are not efficient, and the existing attention mechanisms have high computational complexity and high GPU occupancy. Findings Experimental evaluation using ten open-source Java data sets from PROMISE and five existing methods demonstrates that the proposed approach achieves an F-measure value of 0.637 in predicting 16 cross-version projects, representing a 6.1% improvement. Originality/value RCCA-WRSR is a new visual software defect prediction based on recurrent criss-cross attention and improved residual network. This method effectively enhances the performance of software defect prediction.