International Journal of Innovative Computing and Applications (5 papers in press)

Regular Issues

• Comprehensive Analytical Research on Highly Similar Immunoassay Data based on ResNet and Convolutional Neural Networks  
  by Jianzhang Li, Zixuan Zhao 
  Abstract: Image recognition has become essential in biomedical research, particularly for disease diagnosis and biomarker detection. While Convolutional Neural Networks (CNNs) have achieved success in image tasks, their performance declines with complex, noisy biomedical data. This study compares ResNet and traditional CNNs in antibody immune detection. ResNet's residual connections effectively address vanishing gradients in deep networks, improving training stability and maintaining consistent performance. Its superior feature extraction captures subtle image variations, achieving high classification accuracy under noisy conditions. ResNet also shows strong robustness across different network depths. Experimental results demonstrate that ResNet outperforms conventional CNNs in detection accuracy, especially with specialised biomedical images, and holds significant promise for clinical application. The findings indicate that ResNet is a more reliable and accurate framework for biomedical image recognition, especially in complex and high-noise environments, offering advantages in both research and practical diagnostic contexts.
  Keywords: ResNet; CNN; immunoassay data; recognition.
  DOI: 10.1504/IJICA.2025.10073443
   
  
• Gait Analysis in the Age of Artificial Intelligence: A Comprehensive Review of Advances, Challenges, and Future Directions  
  by Stobak Dutta, Anirban Mitra, Subrata Paul 
  Abstract: Gait analysis, a key biometric modality, identifies individuals via unique walking patterns and is applied in security, medical diagnostics, and surveillance. Over the years, techniques have evolved from model-based and appearance-based methods to advanced Artificial Intelligence (AI) driven systems. AI integration has significantly enhanced feature extraction, classification accuracy, and robustness under real-world challenges. This review emphasizes AI's role in gait analysis, providing a comprehensive overview of methodologies, key datasets, evaluation metrics, and classification based on body representation, temporal modelling, and sensor modalities. It highlights major AI advancements, including deep learning for cross-view gait recognition, in-the-wild scenarios, and multimodal analysis using depth and infrared sensors. The study also examines challenges such as gait variability, occlusion, and biometric privacy concerns. By synthesising existing research and identifying current gaps, this work serves as a valuable reference for researchers and practitioners, and it outlines future directions, emphasising AI's growing influence on gait analysis advancements.
  Keywords: human gait analysis; machine learning algorithms; gait parameters; biomechanics; kinematics; Clinical gait dataset; abnormal gait detection.
  DOI: 10.1504/IJICA.2025.10073597
   
  
• Improvement of Convolutional Neural Networks in Image Classification and Recognition  
  by Pengju Xia 
  Abstract: Image classification, a vital deep learning technique in vision, has been widely applied in fields such as agricultural detection. However, traditional methods often suffer from low efficiency and weak analytical capability. To address this, we propose a CNN-ELM based image classification model incorporating a feature reorganisation attention mechanism and capsule networks for extracting complex semantic information. Experimental results show that when the iteration number of the fusion algorithm reaches 200, the average loss value is 0.026, significantly lower than that of the transfer learning ensemble algorithm (0.082) and the FPGA algorithm (0.087). Moreover, the hybrid model achieves a recognition accuracy of 0.99 for animal-type images, compared to 0.91 with the transfer learning ensemble model. These findings demonstrate that the proposed approach effectively enhances feature recognition and information capture, offering promising potential for structural information analysis and advancing medical imaging applications.
  Keywords: Convolutional neural network; Image classification and recognition; Feature reorganization attention mechanism; Capsule network; Extreme learning machine.
  DOI: 10.1504/IJICA.2025.10073608
   
  
• Enhancing Digital Art Style Recognition via a Hybrid Vision Transformer and Lightweight CNN with Attention Mechanisms  
  by Ying Zhang, Fangzheng Lv 
  Abstract: Current methods for art style recognition often struggle to capture local details and balance global and texture features, leading to vague style representation during multi-scale fusion. To address this, a hybrid model based on Vision Transformer is proposed. It integrates a lightweight CNN branch enhanced with channel attention and dynamic convolution, along with a multi-scale attention-weighted fusion strategy. Experimental results show the model achieves a Kappa coefficient of 0.96, outperforming comparison models (0.78 and 0.74). The structural similarity index reaches 0.94, indicating high image quality and structural fidelity. A no-reference image quality assessment yields a low score of 19.7 after 100 iterations, demonstrating excellent generation performance. The proposed model significantly improves accuracy and feature representation in digital art style recognition, supporting the intelligent development of the cultural and creative industry and advancing deep learning applications in art analysis.
  Keywords: Vision transformer; Lightweight convolutional neural network; Attention mechanism; Digital art; Style recognition.
  DOI: 10.1504/IJICA.2025.10073624
   
  
• Fuzzy Set Induced Coevolutionary Approach to many Objective Optimisation  
  by Selina Khoirom, Pratyusha Rakshit 
  Abstract: The research introduces a new many objective optimization (MaOO) technique that utilizes the parallel nature of evolutionary processes through coevolution. This method applies the artificial bee colony (ABC) algorithm to tackle individual objectives simultaneously resembling coevolution. Once convergence is achieved, a selection of top-quality solutions is carefully chosen from each ABC population, each concentrating on a specific objective. These selections are then combined into a single set. Following this, a ranking system based on fuzzy membership is created to identify the best-performing elements within this combined set, representing the approximate Pareto optimal solutions for the given MaOO problem. The proposed algorithm, called coevolutionary fuzzy-bee colony (CFBC), is tested against three advanced techniques. The experimental outcomes show that CFBC outperforms its competitors in terms of performance metrics. This innovative MaOO approach provides a promising method for optimizing complex problems with multiple objectives, potentially improving decision-making processes in various fields of application.
  Keywords: artificial bee colony; many-objective optimization; parallel optimization; fuzzy membership; multiple population.
  DOI: 10.1504/IJICA.2025.10074047