Propose an architectural change to a pre-trained CNN to better replicate human visual perception in high-resolution image classification tasks. Explain the tradeoffs involved in your design choice.

Modify the CNN architecture to incorporate a multi-scale feature representation, allowing it to capture and process small regions of interest within larger images. This can be achieved by using spatial pyramid pooling or other techniques that enable the model to stitch together multiple glimpses of small areas, mimicking human visual perception at the cost of increased computational complexity and memory requirements.