# 3D Prior is All You Need: Cross-Task Few-shot 2D Gaze Estimation

arXiv: https://arxiv.org/abs/2502.04074 · HTML: https://mlsanity.com/paper/2502.04074/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 22 of its 42 references that are in the map.
- First posted on arXiv 6 Feb 2025.

## Builds on (most cited first, 22 of 22)

- [Towards Fast, Accurate and Stable 3D Dense Face Alignment](https://mlsanity.com/paper/2009.09960/) (2020)
- [Gaze360: Physically Unconstrained Gaze Estimation in the Wild](https://mlsanity.com/paper/1910.10088/) (2019)
- [Eye Tracking for Everyone](https://mlsanity.com/paper/1606.05814/) (2016)
- [Appearance-Based Gaze Estimation in the Wild](https://mlsanity.com/paper/1504.02863/) (2015)
- [It's Written All Over Your Face: Full-Face Appearance-Based Gaze Estimation](https://mlsanity.com/paper/1611.08860/) (2016)
- [Appearance-based Gaze Estimation With Deep Learning: A Review and Benchmark](https://mlsanity.com/paper/2104.12668/) (2021)
- [Predicting the Driver's Focus of Attention: the DR(eye)VE Project](https://mlsanity.com/paper/1705.03854/) (2017)
- [Few-Shot Adaptive Gaze Estimation](https://mlsanity.com/paper/1905.01941/) (2019)
- [Gaze Estimation using Transformer](https://mlsanity.com/paper/2105.14424/) (2021)
- [A Coarse-to-Fine Adaptive Network for Appearance-Based Gaze Estimation](https://mlsanity.com/paper/2001.00187/) (2020)
- [PureGaze: Purifying Gaze Feature for Generalizable Gaze Estimation](https://mlsanity.com/paper/2103.13173/) (2021)
- [Towards End-to-end Video-based Eye-Tracking](https://mlsanity.com/paper/2007.13120/) (2020)
- [Follow My Eye: Using Gaze to Supervise Computer-Aided Diagnosis](https://mlsanity.com/paper/2204.02976/) (2022)
- [Evaluation of Appearance-Based Methods and Implications for Gaze-Based Applications](https://mlsanity.com/paper/1901.10906/) (2019)
- [Adaptive Feature Fusion Network for Gaze Tracking in Mobile Tablets](https://mlsanity.com/paper/2103.11119/) (2021)
- [Photo-Realistic Monocular Gaze Redirection Using Generative Adversarial Networks](https://mlsanity.com/paper/1903.12530/) (2019)
- [DVGaze: Dual-View Gaze Estimation](https://mlsanity.com/paper/2308.10310/) (2023)
- [What Do You See in Vehicle? Comprehensive Vision Solution for In-Vehicle Gaze Estimation](https://mlsanity.com/paper/2403.15664/) (2024)
- [Spatio-Temporal Attention and Gaussian Processes for Personalized Video Gaze Estimation](https://mlsanity.com/paper/2404.05215/) (2024)
- [EFE: End-to-end Frame-to-Gaze Estimation](https://mlsanity.com/paper/2305.05526/) (2023)
- [LG-Gaze: Learning Geometry-aware Continuous Prompts for Language-Guided Gaze Estimation](https://mlsanity.com/paper/2411.08606/) (2024)
- [Efficiency in Real-time Webcam Gaze Tracking](https://mlsanity.com/paper/2009.01270/) (2020)

## Built on by (most cited first, 3 of 3)

- [Real-time Appearance-based Gaze Estimation for Open Domains](https://mlsanity.com/paper/2603.26945/) (2026)
- [LISA: Language-guided Interference-aware Spatial-Frequency Attention for Driver Gaze Estimation](https://mlsanity.com/paper/2605.17287/) (2026)
- [Factor-Informed Uncertainty Distillation for Gaze Estimation](https://mlsanity.com/paper/2607.20072/) (2026)

## Abstract (arXiv)

3D and 2D gaze estimation share the fundamental objective of capturing eye movements but are traditionally treated as two distinct research domains. In this paper, we introduce a novel cross-task few-shot 2D gaze estimation approach, aiming to adapt a pre-trained 3D gaze estimation network for 2D gaze prediction on unseen devices using only a few training images. This task is highly challenging due to the domain gap between 3D and 2D gaze, unknown screen poses, and limited training data. To address these challenges, we propose a novel framework that bridges the gap between 3D and 2D gaze. Our framework contains a physics-based differentiable projection module with learnable parameters to model screen poses and project 3D gaze into 2D gaze. The framework is fully differentiable and can integrate into existing 3D gaze networks without modifying their original architecture. Additionally, we introduce a dynamic pseudo-labelling strategy for flipped images, which is particularly challenging for 2D labels due to unknown screen poses. To overcome this, we reverse the projection process by converting 2D labels to 3D space, where flipping is performed. Notably, this 3D space is not aligned with the camera coordinate system, so we learn a dynamic transformation matrix to compensate for this misalignment. We evaluate our method on MPIIGaze, EVE, and GazeCapture datasets, collected respectively on laptops, desktop computers, and mobile devices. The superior performance highlights the effectiveness of our approach, and demonstrates its strong potential for real-world applications.