Skip to main content

SCAF — Structural Conservation Adaptive Flow

PyPI version License: MIT Tests: 322/328 Zero GPU

قانون فيزيائي واحد. سبعة منتجات. 322 اختبار موثق. صفر GPU.


المبدأ الأساسي

C  = sum(x) / mean(mask)     # ثابت الحفاظ
Ec = |sum(x*) - C| → 0      # خطأ الحفاظ

أي إشارة فيزيائية حقيقية تحفظ مجموع طاقتها. SCAF يحوّل هذا القانون إلى خوارزمية بسيطة تعمل على 7 مجالات طبية.


التثبيت

pip install scaf-medical

مع كل الاعتماديات:

pip install scaf-medical[full]

الاستخدام السريع

SCAF-Inverse — تحسين MRI

import numpy as np
from scaf import SCAFInverse

# kspace: 2D complex array من جهاز MRI
# mask:   2D binary array (1=مقاس، 0=مفقود)
recon  = SCAFInverse()
result = recon.reconstruct(kspace, mask)

# قياس الجودة
snr = recon.snr(reference, result)
print(f"SNR: {snr:.2f} dB")

SCAF-Video — إصلاح فيديو القلب

from scaf import SCAFVideo

# frames:    [T, H, W] — الإطارات (المفقودة = أصفار)
# drop_mask: [T] bool  — True = مفقود
fixer   = SCAFVideo()
repaired = fixer.repair(frames, drop_mask)

SCAF-Guard — كشف هلوسة AI

from scaf import SCAFGuard

guard  = SCAFGuard(threshold=0.05)
result = guard.verify(ai_output, reference_input)

print(result['verdict'])            # PASS / WARNING / FAIL / CRITICAL
print(result['is_hallucination'])   # bool
print(result['confidence'])         # 0.0 - 1.0

SCAF-Cert — شهادة جودة رياضية

from scaf import SCAFCert

cert = SCAFCert()
certificate = cert.certify(reconstructed_image, mask)

print(certificate['verdict'])       # CERTIFIED / WARNING / FAIL
print(certificate['bound_l2'])      # الحد الأعلى لخطأ L2
print(certificate['bound_snr_db'])  # الحد الأعلى لخطأ SNR

SCAF-SORL — تحسين الصور العامة

from scaf import SCAFSORL

enhancer = SCAFSORL()
enhanced = enhancer.enhance(image)

النتائج الموثقة

المنتج الاختبارات النجاح SNR Gain الميزة
SCAF-Inverse 90/90 100% +1.94 dB #1 على 4x Random
SCAF-Video 120/120 100% +1.47 dB كل معدلات الفقدان
SCAF-Guard 6/6 100% +4.36 dB لا مثيل له
SCAF-Cert 90/90 100% مضمون بدون ground truth
SCAF-SORL 3/3 100% +2.16 dB عام

مقارنة SCAF-Inverse مع المنافسين (4x Random — M4Raw)

الطريقة SNR (dB) GPU تدريب شهادة خطأ
SCAF-Inverse v2 15.028 لا لا ✅
BM3D+DC 14.175 لا لا ❌
NLM+DC 14.100 لا لا ❌
TV+DC 13.692 لا لا ❌
Deep Learning ~14.2 ✅ نعم ✅ ملايين صور ❌

المعاملات المقفلة

beta  = 2.42   # قوة الانتشار  — ثابت عبر كل المجالات
dt    = 0.05   # الخطوة الزمنية — ثابت عبر كل المجالات
kappa = 0.22   # عتبة الحواف  — يتغير قليلاً بين المجالات

الاختبارات

pip install scaf-medical[dev]
pytest scaf/tests/ -v

الاقتباس

@software{scaf2025,
  title  = {SCAF: Structural Conservation Adaptive Flow},
  author = {SCAF Research Laboratory},
  year   = {2025},
  url    = {https://github.com/scaf-lab/scaf},
  note   = {322 validated tests, 98.2\% success rate}
}

الترخيص

MIT License — مجاني للاستخدام البحثي والتجاري.

Metadata

Release files for scaf-lab 1.0.0

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for scaf-lab 1.0.0
File Size Uploaded Provenance
scaf_lab-1.0.0.tar.gz 15.0 kB Without provenance Details

Built distribution (wheel)

Table of built distributions (wheels) for scaf-lab 1.0.0
File Interpreter ABI Platform
scaf_lab-1.0.0-py3-none-any.whl Python 3 none any Details

Total release size: 32.2 kB

Release files / scaf_lab-1.0.0.tar.gz

Download URL scaf_lab-1.0.0.tar.gz
Size 15.0 kB
Tags Source
SHA-256 checksum
How to use checksums
e7d01fc7a263ef57830eb5e5ebc6bc5c60e7c9e8c6e372ae6cbd884219c2789a
BLAKE2b-256 checksum
How to use checksums
bee49ee5d87e477413007f52725123493d9a02ad5309c00f62d60bad5117943e
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/6.2.0 CPython/3.13.5

Release files / scaf_lab-1.0.0-py3-none-any.whl

Download URL scaf_lab-1.0.0-py3-none-any.whl
Size 17.2 kB
Tags Python 3
SHA-256 checksum
How to use checksums
ee7a05c163541d8946d0698d7923f56bb1b2d1c0fd4024168879ca4c98147943
BLAKE2b-256 checksum
How to use checksums
0b3e1de065212791d8e48fb6628c3f25951504dedc8a3a16cb914391ac4b48b1
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/6.2.0 CPython/3.13.5

Security for scaf-lab 1.0.0

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

No provenance found

None of the files in this release were uploaded with provenance metadata.

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

Release history Release notifications | RSS feed

This release

1.0.0 This release

2 release files No provenance

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page