# VGGT-HPE: Reframing Head Pose Estimation as Relative Pose Prediction

arXiv: https://arxiv.org/abs/2604.10106 · HTML: https://mlsanity.com/paper/2604.10106/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 21 of its 45 references that are in the map.
- First posted on arXiv 11 Apr 2026.

## Builds on (most cited first, 21 of 21)

- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [DUSt3R: Geometric 3D Vision Made Easy](https://mlsanity.com/paper/2312.14132/) (2023)
- [VGGT: Visual Geometry Grounded Transformer](https://mlsanity.com/paper/2503.11651/) (2025)
- [Joint Face Detection and Alignment using Multi-task Cascaded Convolutional Networks](https://mlsanity.com/paper/1604.02878/) (2016)
- [DROID-SLAM: Deep Visual SLAM for Monocular, Stereo, and RGB-D Cameras](https://mlsanity.com/paper/2108.10869/) (2021)
- [Affect Analysis in-the-wild: Valence-Arousal, Expressions, Action Units and a Unified Framework](https://mlsanity.com/paper/2103.15792/) (2021)
- [Fine-Grained Head Pose Estimation Without Keypoints](https://mlsanity.com/paper/1710.00925/) (2017)
- [MPIIGaze: Real-World Dataset and Deep Appearance-Based Gaze Estimation](https://mlsanity.com/paper/1711.09017/) (2017)
- [TartanVO: A Generalizable Learning-based VO](https://mlsanity.com/paper/2011.00359/) (2020)
- [6D Rotation Representation For Unconstrained Head Pose Estimation](https://mlsanity.com/paper/2202.12555/) (2022)
- [img2pose: Face Alignment and Detection via 6DoF, Face Pose Estimation](https://mlsanity.com/paper/2012.07791/) (2020)
- [Quantitative Survey of the State of the Art in Sign Language Recognition](https://mlsanity.com/paper/2008.09918/) (2020)
- [WHENet: Real-time Fine-Grained Estimation for Wide Range Head Pose](https://mlsanity.com/paper/2005.10353/) (2020)
- [A Vector-based Representation to Enhance Head Pose Estimation](https://mlsanity.com/paper/2010.07184/) (2020)
- [HAAR: Text-Conditioned Generative Model of 3D Strand-based Human Hairstyles](https://mlsanity.com/paper/2312.11666/) (2023)
- [Multi-task head pose estimation in-the-wild](https://mlsanity.com/paper/2202.02299/) (2022)
- [A Review of Verbal and Non-Verbal Human-Robot Interactive Communication](https://mlsanity.com/paper/1401.4994/) (2014)
- [On the representation and methodology for wide and short range head pose estimation](https://mlsanity.com/paper/2401.05807/) (2024)
- [Facial Pose Estimation by Deep Learning from Label Distributions](https://mlsanity.com/paper/1904.13102/) (2019)
- [Towards 3D Face Reconstruction in Perspective Projection: Estimating 6DoF Face Pose from Monocular Image](https://mlsanity.com/paper/2205.04126/) (2022)
- [Towards Unbiased Label Distribution Learning for Facial Pose Estimation Using Anisotropic Spherical Gaussian](https://mlsanity.com/paper/2208.09122/) (2022)

## Abstract (arXiv)

Monocular head pose estimation is traditionally formulated as direct regression from a single image to an absolute pose. This paradigm forces the network to implicitly internalize a dataset-specific canonical reference frame. In this work, we argue that predicting the relative rigid transformation between two observed head configurations is a fundamentally easier and more robust formulation. We introduce VGGT-HPE, a relative head pose estimator built upon a general-purpose geometry foundation model. Finetuned exclusively on synthetic facial renderings, our method sidesteps the need for an implicit anchor by reducing the problem to estimating a geometric displacement from an explicitly provided anchor with a known pose. As a practical benefit, the relative formulation also allows the anchor to be chosen at test time - for instance, a near-neutral frame or a temporally adjacent one - so that the prediction difficulty can be controlled by the application. Despite zero real-world training data, VGGT-HPE achieves state-of-the-art results on the BIWI benchmark, outperforming established absolute regression methods trained on mixed and real datasets. Through controlled easy- and hard-pair benchmarks, we also systematically validate our core hypothesis: relative prediction is intrinsically more accurate than absolute regression, with the advantage scaling alongside the difficulty of the target pose. Project page and code: https://vasilikivas.github.io/VGGT-HPE