# How far are we from solving the 2D & 3D Face Alignment problem? (and a dataset of 230,000 3D facial landmarks)

## TL;DR

This paper evaluates the current state of 2D and 3D face alignment using deep neural networks, introduces a large 3D landmark dataset LS3D-W, and demonstrates that existing models are nearing dataset performance saturation.

## Contribution

It constructs a strong baseline for face alignment, creates the LS3D-W dataset, and analyzes factors affecting alignment performance, highlighting near-saturation of current methods.

## Key findings

- State-of-the-art models nearly saturate dataset performance.
- Introduced LS3D-W, the largest 3D facial landmark dataset.
- Analyzed factors like pose, initialization, resolution, and network size.

## Abstract

This paper investigates how far a very deep neural network is from attaining close to saturating performance on existing 2D and 3D face alignment datasets. To this end, we make the following 5 contributions: (a) we construct, for the first time, a very strong baseline by combining a state-of-the-art architecture for landmark localization with a state-of-the-art residual block, train it on a very large yet synthetically expanded 2D facial landmark dataset and finally evaluate it on all other 2D facial landmark datasets. (b) We create a guided by 2D landmarks network which converts 2D landmark annotations to 3D and unifies all existing datasets, leading to the creation of LS3D-W, the largest and most challenging 3D facial landmark dataset to date ~230,000 images. (c) Following that, we train a neural network for 3D face alignment and evaluate it on the newly introduced LS3D-W. (d) We further look into the effect of all "traditional" factors affecting face alignment performance like large pose, initialization and resolution, and introduce a "new" one, namely the size of the network. (e) We show that both 2D and 3D face alignment networks achieve performance of remarkable accuracy which is probably close to saturating the datasets used. Training and testing code as well as the dataset can be downloaded from https://www.adrianbulat.com/face-alignment/