# GADS: A Super Lightweight Model for Head Pose Estimation

arXiv: https://arxiv.org/abs/2504.15751 · HTML: https://mlsanity.com/paper/2504.15751/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 24 of its 51 references that are in the map.
- First posted on arXiv 22 Apr 2025.

## Builds on (most cited first, 24 of 24)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation](https://mlsanity.com/paper/1612.00593/) (2016)
- [Neural Machine Translation by Jointly Learning to Align and Translate](https://mlsanity.com/paper/1409.0473/) (2014)
- [PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space](https://mlsanity.com/paper/1706.02413/) (2017)
- [Deep Sets](https://mlsanity.com/paper/1703.06114/) (2017)
- [Self-Attention with Relative Position Representations](https://mlsanity.com/paper/1803.02155/) (2018)
- [Joint Face Detection and Alignment using Multi-task Cascaded Convolutional Networks](https://mlsanity.com/paper/1604.02878/) (2016)
- [MediaPipe: A Framework for Building Perception Pipelines](https://mlsanity.com/paper/1906.08172/) (2019)
- [Deep Learning for 3D Point Clouds: A Survey](https://mlsanity.com/paper/1912.12033/) (2019)
- [How far are we from solving the 2D & 3D Face Alignment problem? (and a dataset of 230,000 3D facial landmarks)](https://mlsanity.com/paper/1703.07332/) (2017)
- [Attention-Based Models for Speech Recognition](https://mlsanity.com/paper/1506.07503/) (2015)
- [Recent Advances in Convolutional Neural Networks](https://mlsanity.com/paper/1512.07108/) (2015)
- [Fine-Grained Head Pose Estimation Without Keypoints](https://mlsanity.com/paper/1710.00925/) (2017)
- [Face Alignment in Full Pose Range: A 3D Total Solution](https://mlsanity.com/paper/1804.01005/) (2018)
- [img2pose: Face Alignment and Detection via 6DoF, Face Pose Estimation](https://mlsanity.com/paper/2012.07791/) (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)
- [Every Smile is Unique: Landmark-Guided Diverse Smile Generation](https://mlsanity.com/paper/1802.01873/) (2018)
- [POSEidon: Face-from-Depth for Driver Pose Estimation](https://mlsanity.com/paper/1611.10195/) (2016)
- [Light-weight Head Pose Invariant Gaze Tracking](https://mlsanity.com/paper/1804.08572/) (2018)
- [KEPLER: Keypoint and Pose Estimation of Unconstrained Faces by Learning Efficient H-CNN Regressors](https://mlsanity.com/paper/1702.05085/) (2017)
- [Improving Head Pose Estimation with a Combined Loss and Bounding Box Margin Adjustment](https://mlsanity.com/paper/1905.08609/) (2019)
- [Face-from-Depth for Head Pose Estimation on Depth Images](https://mlsanity.com/paper/1712.05277/) (2017)

## Abstract (arXiv)

In human-computer interaction, head pose estimation profoundly influences application functionality. Although utilizing facial landmarks is valuable for this purpose, existing landmark-based methods prioritize precision over simplicity and model size, limiting their deployment on edge devices and in compute-poor environments. To bridge this gap, we propose Grouped Attention Deep Sets (GADS), a novel architecture based on the Deep Set framework. By grouping landmarks into regions and employing small Deep Set layers, we reduce computational complexity. Our multihead attention mechanism extracts and combines inter-group information, resulting in a model that is 7.5× smaller and executes 25× faster than the current lightest state-of-the-art model. Notably, our method achieves an impressive reduction, being 4321× smaller than the best-performing model. We introduce vanilla GADS and Hybrid-GADS (landmarks + RGB) and evaluate our models on three benchmark datasets -- AFLW2000, BIWI, and 300W-LP. We envision our architecture as a robust baseline for resource-constrained head pose estimation methods.