# DCI: Discriminative and Contrast Invertible Descriptor

## TL;DR

The paper introduces DCI, a new local feature descriptor designed to improve robustness under illumination changes by using Laplace gradient histograms and contrast flip estimation, enhancing fine-grained object recognition.

## Contribution

It presents a novel discriminative and contrast invertible descriptor that outperforms existing descriptors in challenging illumination conditions.

## Key findings

- DCI outperforms existing descriptors in recognition tasks.
- The Laplace gradient histogram enhances discriminative power.
- Contrast flip estimation improves robustness to illumination changes.

## Abstract

Local feature descriptors have been widely used in fine-grained visual object search thanks to their robustness in scale and rotation variation and cluttered background. However, the performance of such descriptors drops under severe illumination changes. In this paper, we proposed a Discriminative and Contrast Invertible (DCI) local feature descriptor. In order to increase the discriminative ability of the descriptor under illumination changes, a Laplace gradient based histogram is proposed. A robust contrast flipping estimate is proposed based on the divergence of a local region. Experiments on fine-grained object recognition and retrieval applications demonstrate the superior performance of DCI descriptor to others.