Used for image recognition tasks?

• A. Convolutional neural networks (CNNs)
• B. Generative AI
• C. Hybrid approaches
• D. Mask learning

Answer: (A)

Image recognition relies on understanding spatial relationships in pixel data. Convolutional neural networks are built for grid-structured inputs like images: they use small, learnable filters that slide across the image to detect simple patterns such as edges, then stack many layers to build up increasingly complex features. The weight-sharing and local connectivity mean the same detector works across the image, and pooling helps reduce dimensionality while preserving essential information, giving translation invariance. This combination enables robust recognition even in large, diverse datasets. Generative AI focuses on creating new images or content rather than classification, hybrid approaches mix methods but the most direct, effective mechanism for recognizing visual content is provided by CNNs, and mask learning typically pertains to segmentation tasks rather than general recognition.