Pixel-level feature attributions are an important tool in eXplainable AI for Computer Vision (XCV), providing visual insights into how image features influence model predictions. The Owen formula for hierarchical Shapley values has been widely used to interpret machine learning (ML) models and their learned representations. However, existing hierarchical Shapley approaches do not exploit the multiscale structure of image data, leading to slow convergence and weak alignment with the actual morphological features. Moreover, no prior Shapley method has leveraged data-aware hierarchies for Computer Vision tasks, leaving a gap in model interpretability of structured visual data. To address this, this paper introduces ShapBPT, a novel data-aware XCV method based on the hierarchical Shapley formula. ShapBPT assigns Shapley coefficients to a multiscale hierarchical structure tailored for images, the Binary Partition Tree (BPT). By using this data-aware hierarchical partitioning, ShapBPT ensures that feature attributions align with intrinsic image morphology, effectively prioritizing relevant regions while reducing computational overhead. This advancement connects hierarchical Shapley methods with image data, providing a more efficient and semantically meaningful approach to visual interpretability. Experimental results confirm ShapBPT’s effectiveness, demonstrating superior alignment with image structures and improved efficiency over existing XCV methods, and a 20-subject user study confirming that ShapBPT explanations are preferred by humans.

We investigate the use of a stratified sampling approach for LIME Image, a popular model-agnostic explainable AI method for computer vision tasks, in order to reduce the artifacts generated by typical Monte Carlo sampling. Such artifacts are due to the undersampling of the dependent variable in the synthetic neighborhood around the image being explained, which may result in inadequate explanations due to the impossibility of fitting a linear regressor on the sampled data. We then highlight a connection with the Shapley theory, where similar arguments about undersampling and sample relevance were suggested in the past. We derive all the formulas and adjustment factors required for an unbiased stratified sampling estimator. Experiments show the efficacy of the proposed approach.

Gastrointestinal tract (GIT) infections such as ulcers, bleeding, polyps, Crohn’s disease and cancer are quite familiar today worldwide. Wireless capsule endoscopy (WCE) is an efficient means of investigation of GIT diseases. However, still several challenges exist in this domain, such as lesion shape, colour, texture, size and irregularity. To deal with these problems, several computer-based methods are introduced in computer vision domain but they used only hand-crafted features which produced wrong predictions several times. In this research, a new technique is applied based on the fusion of deep convolutional (CNN) and geometric features. Initially, disease regions are extracted from given WCE images using a new approach named contrast-enhanced colour features . Geometric features are extracted from segmented disease region. Thereafter, unique VGG16 and VGG19 deep CNN features fusion are performed based on Euclidean Fisher Vector . The unique features are fused with geometric features which are later fed to conditional entropy approach for best features selection. The selected features are finally classified by K-Nearest Neighbour . A privately collected database which consists of 5500 WCE images is utilised for the evaluation of the proposed method and achieved best classification accuracy of 99.42% and precision rate of 99.51%. The classification accuracy proves the authenticity of the proposed approach.