Metadata-Version: 2.5
Name: seetapsych-gaze
Version: 0.0.2
Summary: Gaze estimation modules for SeetaPsych
Project-URL: Homepage, https://github.com/seetapsych/seetapsych-gaze
Project-URL: Repository, https://github.com/seetapsych/seetapsych-gaze
Project-URL: Issues, https://github.com/seetapsych/seetapsych-gaze/issues
License: Copyright (c) 2026, Visual Information Processing and Learning (VIPL) group,
        Institute of Computing Technology, Chinese Academy of Sciences, Beijing, China;
        Southeast University, China;
        Beijing Seetatech Co., Ltd.
        All rights reserved.
        
        Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:
        
        1. Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.
        
        2. Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.
        
        3. Neither the name of the copyright holder nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission.
        
        THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
License-File: LICENSE
Requires-Python: >=3.10
Requires-Dist: seetapsych-lib
Provides-Extra: all
Requires-Dist: mediapipe>=0.10.33; extra == 'all'
Requires-Dist: mmengine>=0.10.7; extra == 'all'
Requires-Dist: numpy>=2.2.6; extra == 'all'
Requires-Dist: onnx>=1.20.1; extra == 'all'
Requires-Dist: onnxruntime-gpu<1.24; (sys_platform != 'darwin' and python_version <= '3.10') and extra == 'all'
Requires-Dist: onnxruntime-gpu>=1.24; (sys_platform != 'darwin' and python_version > '3.10') and extra == 'all'
Requires-Dist: onnxruntime<1.24; (sys_platform == 'darwin' and python_version <= '3.10') and extra == 'all'
Requires-Dist: onnxruntime>=1.24; (sys_platform == 'darwin' and python_version > '3.10') and extra == 'all'
Requires-Dist: opencv-python>=4.13.0.92; extra == 'all'
Requires-Dist: pillow>=12.2.0; extra == 'all'
Requires-Dist: safetensors>=0.7.0; extra == 'all'
Requires-Dist: scipy>=1.15.3; extra == 'all'
Requires-Dist: timm>=1.0.26; extra == 'all'
Requires-Dist: torch>=2.10.0; extra == 'all'
Requires-Dist: torchvision>=0.25.0; extra == 'all'
Provides-Extra: dev
Requires-Dist: bandit>=1.7; extra == 'dev'
Requires-Dist: build>=1.4.4; extra == 'dev'
Requires-Dist: mediapipe>=0.10.33; extra == 'dev'
Requires-Dist: mmengine>=0.10.7; extra == 'dev'
Requires-Dist: mypy>=1.11; extra == 'dev'
Requires-Dist: numpy>=2.2.6; extra == 'dev'
Requires-Dist: onnx>=1.20.1; extra == 'dev'
Requires-Dist: onnxruntime-gpu<1.24; (sys_platform != 'darwin' and python_version <= '3.10') and extra == 'dev'
Requires-Dist: onnxruntime-gpu>=1.24; (sys_platform != 'darwin' and python_version > '3.10') and extra == 'dev'
Requires-Dist: onnxruntime<1.24; (sys_platform == 'darwin' and python_version <= '3.10') and extra == 'dev'
Requires-Dist: onnxruntime>=1.24; (sys_platform == 'darwin' and python_version > '3.10') and extra == 'dev'
Requires-Dist: opencv-python>=4.13.0.92; extra == 'dev'
Requires-Dist: pillow>=12.2.0; extra == 'dev'
Requires-Dist: pytest>=8.0; extra == 'dev'
Requires-Dist: ruff>=0.6; extra == 'dev'
Requires-Dist: safetensors>=0.7.0; extra == 'dev'
Requires-Dist: scipy-stubs>=1.15.3.0; extra == 'dev'
Requires-Dist: scipy>=1.15.3; extra == 'dev'
Requires-Dist: timm>=1.0.26; extra == 'dev'
Requires-Dist: torch>=2.10.0; extra == 'dev'
Requires-Dist: torchvision>=0.25.0; extra == 'dev'
Requires-Dist: twine>=6.2.0; extra == 'dev'
Description-Content-Type: text/markdown

# SeetaPsych Gaze

> Gaze estimation modules for SeetaPsych

## Usage

This project is already included in the seetapsych-lib default configuration. Download and use it via `seetapsych-manager download`.

For usage, refer to [SeetaPsych](https://github.com/seetapsych/seetapsych-lib).

The gaze estimation algorithms depend on `open-gaze-estimation`. Please install it from GitHub separately:

```bash
uv pip install git+https://github.com/Elorfiniel/open-gaze-estimation-2025-release.git
```

You can additionally add this algorithm module using the following methods.

### WebUI

Run `seetapsych-webui` with the `--files` argument to use it.

```
seetapsych-webui --files \
  seetapsych_gaze/modules/affnet.yml \
  seetapsych_gaze/modules/itracker-plus.yml \
  seetapsych_gaze/modules/tdgazenet.yml
```

### Programmatic Usage

Add the following code in your program to use this algorithm module.

```python
from seetapsych_lib.runtime.factory import Factory
from seetapsych_lib.runtime.pipeline import Pipeline

factory = Factory()
factory.load_file_modules("seetapsych_gaze/modules/affnet.yml")

pipeline = Pipeline(factory, ...)

pipeline.add_attributes("face/gaze-screen")
```

## Introduction

### OpenGaze-AFFNet

PyTorch-based appearance feature fusion gaze estimation model (AFFNet). Accepts face image, left/right eye crops, and face bounding box rectangles as input. Estimates 2D gaze point on screen. Requires `face/mesh` (MediaPipe 468 landmarks) for eye alignment and face cropping.

Module config: [affnet.yml](seetapsych_gaze/modules/affnet.yml).
Provide Attributes: `face/gaze-screen`.

Requires: `face/mesh`.

Available model: `model-AFFNet-mit-gaze-capture-e20.safetensors`.

Parameters:
- `data` (object): Camera and screen calibration settings including intrinsic/extrinsic matrices, screen dimensions.

### OpenGaze-ITrackerPlus

ONNX-based ITrackerPlus gaze estimation model. Accepts face crop, left/right eye crops, and facial keypoints as input. Uses MediaPipe FaceMesh for landmark detection. Estimates 2D gaze point on screen.

Module config: [itracker-plus.yml](seetapsych_gaze/modules/itracker-plus.yml).
Provide Attributes: `face/gaze-screen`.

Requires: `face/mesh`.

Available model: `model-ITrackerPlus-oppo-data-e12.onnx`.

Parameters:
- `data` (object): Camera and screen calibration settings.

### OpenGaze-TdGazeNet

PyTorch-based TdGazeNet (multi-task) gaze estimation model. Predicts 3D face keypoints (151 pts), eye keypoints (110 pts each), and 3D gaze vectors simultaneously. Supports reparameterization optimization for faster inference.

Module config: [tdgazenet.yml](seetapsych_gaze/modules/tdgazenet.yml).
Provide Attributes: `face/gaze-screen`.

Requires: `face/mesh`.

Available model: `model-TdGazeNet-ucas-synthgaze-e50.safetensors`.

Parameters:
- `optimize` (selection, default `none`): `none` or `reparameterize`. Optionally reparameterize the model for inference speedup.
- `data` (object): Camera intrinsic, extrinsic, distortion, screen settings, and model architecture config.
