SeetaPsych Hertz

Heart rate 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.

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

Heart rate estimation requires processing video or real-time video streams to extract heart rate information.

WebUI

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

seetapsych-webui --files seetapsych_hertz/modules/seeta.yml

Programmatic Usage

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

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

factory = Factory()
factory.load_file_modules("seetapsych_hertz/modules/seeta.yml")

pipeline = Pipeline(factory, ...)

pipeline.add_attributes("face/heart_rate")

Module Catalog

Module YAML Path	Package Name
`seetapsych_hertz/modules/ada-chrom.yml`	HeartRate-AdaChrom
`seetapsych_hertz/modules/seeta.yml`	HeartRate-Seeta
`seetapsych_hertz/modules/tiny-hr.yml`	HeartRate-TinyHR

AdaChrom

Model-free rPPG heart rate estimation using adaptive chrominance analysis on skin ROI.

Module config: ada-chrom.yml

Package	Provides	Requires
HeartRate-AdaChrom	`face/heart_rate`	`face/dense_landmarks`

Description

Adaptive chrominance rPPG heart rate estimator. Accepts multiple ROI selectors; the default forehead-only adaptive skin mask (skin_b_adaptive_forehead) matches the original delivery configuration, while the preset group all runs every available region.

Usage Notes

Supports both video streams and video files.
For video stream mode, best results are achieved at 30 FPS or higher, which requires optimized processing logic and better hardware (with GPU).
For stable analysis results, it is recommended to use video files with a stable frame rate of 30 FPS or higher.

Parameters

Name	Type	Default	Description & Tuning
`window_samples`	integer	`300`	Sliding window frame count for HR estimation. Larger values reduce noise but increase latency; adjust based on real-time demand.
`roi_regions`	`selection[]`	`["skin_b_adaptive_forehead"]`	ROI selectors to estimate heart rate on. Multiple selectors are evaluated independently, with valid results merged into the fused `hr_bpm` and the per-region `roi_hr_bpm` map.

Models

(None)

Output Attributes

face/heart_rate — spec.

The per-region results requested via roi_regions are returned inside roi_hr_bpm: each key corresponds to one selected ROI and the value is that region's heart rate in BPM for the current window.

TinyHR

Lightweight neural network for fast heart rate estimation directly from face video frames.

Module config: tiny-hr.yml

Package	Provides	Requires
HeartRate-TinyHR	`face/heart_rate`	`face/detection`

Description

Fast RhythmFormer heart rate estimator using buffered face crops + Welch spectral analysis.

Usage Notes

Supports both video streams and video files.
For video stream processing, frame rates close to 30 or 25 FPS yield the best results.
For stable analysis results, it is recommended to use video files with a frame rate of 30 FPS or 25 FPS.

Parameters

Name	Type	Default	Description & Tuning
`fps`	number	`30`	Expected camera/video FPS used for spectral analysis windowing. Mismatch with actual source FPS degrades HR accuracy.
`interval`	number	`1`	Seconds between consecutive HR estimates. Smaller intervals yield more updates with higher jitter; larger intervals are smoother but slower.

Models

Model	Recommended
`seeta-hertz-tinyhr.onnx`	✓

Output Attributes

face/heart_rate — spec.

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

seetapsych_hertz-0.0.3.post1.tar.gz (208.9 kB view details)

Uploaded Sep 3, 2026 Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

The dropdown lists show the available interpreters, ABIs, and platforms. Enable javascript to be able to filter the list of wheel files.

seetapsych_hertz-0.0.3.post1-py3-none-any.whl (84.2 kB view details)

Uploaded Sep 3, 2026 Python 3

File details

Details for the file seetapsych_hertz-0.0.3.post1.tar.gz.

File metadata

Download URL: seetapsych_hertz-0.0.3.post1.tar.gz
Upload date: Sep 3, 2026
Size: 208.9 kB
Tags: Source
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/7.0.0 CPython/3.13.14

File hashes

Hashes for seetapsych_hertz-0.0.3.post1.tar.gz
Algorithm	Hash digest
SHA256	`5905a8eefe142d3b891c3783b25d002be0bd94b67ba1f78d1a8def51bf90f56d`
MD5	`372d16675349853b9cd7b338c999fa1b`
BLAKE2b-256	`74d5157cddade3ed7cd5b182bd6ed87cbf716e27c9750d65b2234aad1a659d47`

See more details on using hashes here.

Provenance

The following attestation bundles were made for seetapsych_hertz-0.0.3.post1.tar.gz:

Publisher: publish.yml on seetapsych/seetapsych-hertz

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: seetapsych_hertz-0.0.3.post1.tar.gz
- Subject digest: 5905a8eefe142d3b891c3783b25d002be0bd94b67ba1f78d1a8def51bf90f56d
- Sigstore transparency entry: 2698327810
- Sigstore integration time: Sep 3, 2026
Source repository:
- Permalink: seetapsych/seetapsych-hertz@d57f323cb58d2ead397a508b89bb5fe3e547b204
- Branch / Tag: refs/tags/r0.0.3.post1
- Owner: https://github.com/seetapsych
- Access: public
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: publish.yml@d57f323cb58d2ead397a508b89bb5fe3e547b204
- Trigger Event: push