# Learning to Validate the Quality of Detected Landmarks

## TL;DR

This paper introduces a new loss function for CNN-based landmark detection that estimates landmark accuracy, enabling exclusion of unreliable landmarks and improving overall detection quality.

## Contribution

It proposes a novel validation loss function and a batch balancing method that enhances landmark detection accuracy and reliability in CNN models.

## Key findings

- Validation loss correlates with landmark accuracy.
- Batch balancing improves detection performance.
- Method effective across multiple facial landmark datasets.

## Abstract

We present a new loss function for the validation of image landmarks detected via Convolutional Neural Networks (CNN). The network learns to estimate how accurate its landmark estimation is. This loss function is applicable to all regression-based location estimations and allows the exclusion of unreliable landmarks from further processing. In addition, we formulate a novel batch balancing approach which weights the importance of samples based on their produced loss. This is done by computing a probability distribution mapping on an interval from which samples can be selected using a uniform random selection scheme. We conducted experiments on the 300W, AFLW, and WFLW facial landmark datasets. In the first experiments, the influence of our batch balancing approach is evaluated by comparing it against uniform sampling. In addition, we evaluated the impact of the validation loss on the landmark accuracy based on uniform sampling. The last experiments evaluate the correlation of the validation signal with the landmark accuracy. All experiments were performed for all three datasets.