Large Field and High Resolution: Detecting Needle in Haystack
This addresses the computational bottleneck in visual tasks for applications requiring high-resolution analysis over large fields, representing an incremental improvement in efficiency.
The paper tackled the problem of applying CNNs to large-field visual tasks with fine details by developing variable sampling schemes for small target localization, achieving performance that outperforms full-resolution models using only 5% of samples.
The growing use of convolutional neural networks (CNN) for a broad range of visual tasks, including tasks involving fine details, raises the problem of applying such networks to a large field of view, since the amount of computations increases significantly with the number of pixels. To deal effectively with this difficulty, we develop and compare methods of using CNNs for the task of small target localization in natural images, given a limited "budget" of samples to form an image. Inspired in part by human vision, we develop and compare variable sampling schemes, with peak resolution at the center and decreasing resolution with eccentricity, applied iteratively by re-centering the image at the previous predicted target location. The results indicate that variable resolution models significantly outperform constant resolution models. Surprisingly, variable resolution models and in particular multi-channel models, outperform the optimal, "budget-free" full-resolution model, using only 5\% of the samples.