G-CNN: an Iterative Grid Based Object Detector

TL;DR

G-CNN introduces a proposal-free, iterative CNN-based object detection method that starts from a fixed multi-scale grid and refines bounding boxes, achieving comparable accuracy with fewer boxes and faster processing.

Contribution

It presents a novel grid-based iterative detection approach that eliminates the need for proposal algorithms, simplifying and speeding up object detection.

Findings

Achieves comparable accuracy to Fast R-CNN with fewer boxes.

Removes the object proposal stage, reducing computation.

Faster detection process due to fewer boxes and no proposals.

Abstract

We introduce G-CNN, an object detection technique based on CNNs which works without proposal algorithms. G-CNN starts with a multi-scale grid of fixed bounding boxes. We train a regressor to move and scale elements of the grid towards objects iteratively. G-CNN models the problem of object detection as finding a path from a fixed grid to boxes tightly surrounding the objects. G-CNN with around 180 boxes in a multi-scale grid performs comparably to Fast R-CNN which uses around 2K bounding boxes generated with a proposal technique. This strategy makes detection faster by removing the object proposal stage as well as reducing the number of boxes to be processed.