This study presents an advanced modification of the You Only Look Once version 7 (YOLOv7) model named Gallbladder YOLOv7 (GB-YOLOv7). GB-YOLOv7 integrates a Normalization-based Attention Module (NAM) and a Global Attention Mechanism (GAM) into the backbone and head architecture. Several image preprocessing methods are also employed, including median filtering and Contrast-Limited Adaptive Histogram Equalization (CLAHE). The framework includes three attention mechanism-based models, including Coordinate and Global Attention Mechanism (CordGAM-YOLOv7), Dual Global Attention Mechanism (DualGAM-YOLOv7), and Normalization-based Attention Module YOLOv7 (NAM-YOLOv7), enabling a meticulous comparative analysis with GB-YOLOv7. Results demonstrate its superior performance across all metrics compared to both traditional and newer YOLO versions: achieving a Recall of 91.3% (vs YOLOv8's 78.1% and YOLOv9's 84.8%), a Mean Average Precision of 94.0%, and a Specificity of 96.2% (vs YOLOv11's 90.6%). GB-YOLOv7 also shows significant improvements in Matthews Correlation Coefficient (MCC) (72.7% vs YOLOv7's 67.3%) and F1-score (90.0% vs You Only Look Once version 9 ( YOLOv9)'s 81.4%), while maintaining greater parameter efficiency (24.34M vs YOLOv7's 36.58M), showcasing its potential as a cutting-edge tool for more effective gallbladder cancer detection.

Viruses are a significant threat to human life, as demonstrated by the global COVID-19 pandemic and the Ebola outbreak. Diseases such as smallpox, AIDS, hepatitis, liver cancer, and cervical cancer often caused by the Human Papillomavirus can lead to fatal outcomes. Over the years, extensive research has focused on developing vaccines and antiviral drugs, which have successfully contained and, in some cases, eradicated viral infections. Recently, computational techniques, particularly machine learning algorithms, have made notable progress in identifying potential Antiviral Peptides (AVPs), thereby accelerating experimental validation for therapeutic applications. This study introduces machine learning techniques and a Light Gradient Boosting Machine (LGBM) model combined with three feature-encoding methods to predict whether a protein sequence contains effective antiviral peptides. Eight machine learning algorithms were evaluated using both single-feature and combined feature encodings. Among them, the lightweight LGBM model trained on combined encoded features achieved the best performance, with an accuracy of 98%, precision of 97%, recall of 98%, F1-score of 98%, and an AUC of 1.00. Compared to existing models, the proposed approach achieved approximately 2% higher accuracy using individual encoding methods and about 3% higher accuracy with combined features. The reliability and effectiveness of the proposed model highlight its potential value for pharmaceutical development and academic research.

Diagnosing gastrointestinal (GI) disease from endoscopic images remains a challenging task due to high intra-class variability, inter-class similarity, and limited annotated data. To address these issues, this study proposes a novel Dynamic Few-Shot Learning (DFSL) framework that integrates the representational strength of convolutional backbones with a dynamic metric-based classifier to perform accurate classification under low-data regimes. We benchmark our approach using DenseNet-201, ResNet-50, and ViT-B/16 across multiple few-shot settings (1 to 5 shots, up to 4-way classification). Among them, DFSL with ResNet-50 consistently outperforms state-of-the-art backbones, achieving a peak accuracy of 97.6% in the 4-shot scenario. Our experiments reveal that ResNet-50's residual connections enable more robust spatial encoding and generalization across variations in the number of shots. To further enhance interpretability and clinical relevance, Grad-CAM and TCAV were employed to visualize the decision rationale and validate the model’s focus on disease-specific regions. A comparative analysis with supplementary confusion matrices reveals that DFSL-ResNet-50 achieves superior class-wise performance, with reduced misclassification of visually similar diseases. While ViT-B/16 and DenseNet-201 exhibit fluctuating accuracy under scarce data, our model maintains consistent performance. The proposed DFSL framework, with its potential for clinical deployment, presents a significant advancement in developing scalable, interpretable, and accurate diagnostic systems, instilling optimism for its real-world application.

We introduce BioTrove, the largest publicly accessible dataset designed to advance AI applications in biodiversity. Curated from the iNaturalist platform and vetted to include only research-grade data, BioTrove contains 161.9 million images, offering unprecedented scale and diversity from three primary kingdoms: Animalia ("animals"), Fungi ("fungi"), and Plantae ("plants"), spanning approximately 366.6K species. Each image is annotated with scientific names, taxonomic hierarchies, and common names, providing rich metadata to support accurate AI model development across diverse species and ecosystems.We demonstrate the value of BioTrove by releasing a suite of CLIP models trained using a subset of 40 million captioned images, known as BioTrove-Train. This subset focuses on seven categories within the dataset that are underrepresented in standard image recognition models, selected for their critical role in biodiversity and agriculture: Aves ("birds"), Arachnida} ("spiders/ticks/mites"), Insecta ("insects"), Plantae ("plants"), Fungi ("fungi"), Mollusca ("snails"), and Reptilia ("snakes/lizards"). To support rigorous assessment, we introduce several new benchmarks and report model accuracy for zero-shot learning across life stages, rare species, confounding species, and multiple taxonomic levels.We anticipate that BioTrove will spur the development of AI models capable of supporting digital tools for pest control, crop monitoring, biodiversity assessment, and environmental conservation. These advancements are crucial for ensuring food security, preserving ecosystems, and mitigating the impacts of climate change. BioTrove is publicly available, easily accessible, and ready for immediate use.

ABSTRACT Colorectal cancer (CRC) is characterized by the uncontrollable growth of cancerous cells within the rectal mucosa. In contrast, colon polyps, precancerous growths, can develop into colon cancer, causing symptoms like rectal bleeding, abdominal pain, diarrhea, weight loss, and constipation. It is the leading cause of death worldwide, and this potentially fatal cancer severely afflicts the elderly. Furthermore, early diagnosis is crucial for effective treatment, as it is often more time‐consuming and laborious for experts. This study improved the accuracy of CRC multi‐class classification compared to previous research utilizing diverse datasets, such as NCT‐CRC‐HE‐100 K (100,000 images) and CRC‐VAL‐HE‐7 K (7,180 images). Initially, we utilized various image processing techniques on the NCT‐CRC‐HE‐100 K dataset to improve image quality and noise‐freeness, followed by multiple feature extraction and selection methods to identify prominent features from a large data hub and experimenting with different approaches to select the best classifiers for these critical features. The third ensemble model (XGB‐LightGBM‐RF) achieved an optimum accuracy of 99.63% with 40 prominent features using univariate feature selection methods. Moreover, the third ensemble model also achieved 99.73% accuracy from the CRC‐VAL‐HE‐7 K dataset. After combining two datasets, the third ensemble model achieved 99.27% accuracy. In addition, we trained and tested our model with two different datasets. We used 80% data from NCT‐CRC‐HE‐100 K and 20% data from CRC‐VAL‐HE‐7 K, respectively, for training and testing purposes, while the third ensemble model obtained 98.43% accuracy in multi‐class classification. The results show that this new framework, which was created using the third ensemble model, can help experts figure out what kinds of CRC diseases people are dealing with at the very beginning of an investigation.

Abstract The increasing incidence of skin cancer necessitates advancements in early detection methods, where deep learning can be beneficial. This study introduces SkinNet-14, a novel deep learning model designed to classify skin cancer types using low-resolution dermoscopy images. Unlike existing models that require high-resolution images and extensive training times, SkinNet-14 leverages a modified compact convolutional transformer (CCT) architecture to effectively process 32 × 32 pixel images, significantly reducing the computational load and training duration. The framework employs several image preprocessing and augmentation strategies to enhance input image quality and balance the dataset to address class imbalances in medical datasets. The model was tested on three distinct datasets—HAM10000, ISIC and PAD—demonstrating high performance with accuracies of 97.85%, 96.00% and 98.14%, respectively, while significantly reducing the training time to 2–8 s per epoch. Compared to traditional transfer learning models, SkinNet-14 not only improves accuracy but also ensures stability even with smaller training sets. This research addresses a critical gap in automated skin cancer detection, specifically in contexts with limited resources, and highlights the capabilities of transformer-based models that are efficient in medical image analysis.

Recognizing the activities causing distraction in real-world driving scenarios is critical for ensuring the safety and reliability of both drivers and pedestrians on the roadways. Conventional computer vision techniques are typically data-intensive and require a large volume of annotated training data to detect and classify various distracted driving behaviors, thereby limiting their generalization ability, efficiency and scalability. We aim to develop a generalized framework that showcases robust performance with access to limited or no annotated training data. Recently, vision-language models have offered large-scale visual-textual pretraining that can be adapted to task-specific learning like distracted driving activity recognition. Vision-language pretraining models like CLIP have shown significant promise in learning natural language-guided visual representations. This paper proposes a CLIP-based driver activity recognition approach that identifies driver distraction from naturalistic driving images and videos. CLIP’s vision embedding offers zero-shot transfer and task-based finetuning, which can classify distracted activities from naturalistic driving video. Our results show that this framework offers state-of-the-art performance on zero-shot transfer, finetuning and video-based models for predicting the driver’s state on four public datasets. We propose frame-based and video-based frameworks developed on top of the CLIP’s visual representation for distracted driving detection and classification tasks and report the results. Our code is available at https://github.com/zahid-isu/DriveCLIP