# Color Cerberus

## TL;DR

This paper reports that a simple convolutional neural network successfully won a color constancy competition, and suggests that reimplementing a previous neural architecture could improve results further.

## Contribution

Demonstrates the effectiveness of a simple CNN in color constancy and proposes reimplementing a prior architecture for enhanced performance.

## Key findings

- CNN won ISISPA color constancy competition
- Reimplementing Bianco (2017) architecture could improve results
- Simple models can be competitive in color constancy tasks

## Abstract

Simple convolutional neural network was able to win ISISPA color constancy competition. Partial reimplementation of (Bianco, 2017) neural architecture would have shown even better results in this setup.