nunnajaswant-ship-it/skinshield

How nunnajaswant-ship-it/skinshield works

SkinShield is documented as an offline, multimodal skin-triage device, but this repository primarily contains dataset and model research scripts—not an end-to-end device inference application. The documented user workflow combines a lesion image, symptom text, and sensor readings; conditions the inputs; runs a contagion gate before disease identification; and provides referral guidance. The graph also shows the implemented research stages and ONNX export. Device and dashboard capabilities are README-documented; their implementation is not present in the sampled source.

Main entry points: hardware Edge inference runtime · train_dermlip.py DermLIP training · curate_derm1m.py Derm1M curation · export_onnx.py ONNX export

Components 18

Dataset preparation

Derm1M curation codes/curate_derm1m.py
Class verification codes/g1_verify_classes.py

Model research

DermLIP training codes/train_dermlip.py
Contagion gate codes/stage1_gate.py
Gate evaluation codes/stage1_report.py
Disease identification codes/stage2_disease_id.py
Caption de-leaking codes/stage2_deleak.py
Trimodal evaluation codes/stage2plus_trimodal.py
Zero-shot breadth codes/zeroshot_breadth.py
Research queue codes/run_queue.py

Edge deployment

ONNX export codes/export_onnx.py
Edge inference runtime hardware

Triage device

Multimodal input hardware
Sensor dashboard hardware
Triage and referral README.md

Other components

Patient or worker
Derm1M source
Research artifacts

Connections 19

Patient or worker → Multimodal input: provides inputs README.md
Multimodal input → Sensor dashboard: feeds sensor readings README.md
Sensor dashboard → Edge inference runtime: supplies sensor data README.md
Patient or worker → Edge inference runtime: provides image and text README.md
Edge inference runtime → Triage and referral: produces guidance README.md
Derm1M source → Class verification: provides metadata codes/g1_verify_classes.py
Derm1M source → Derm1M curation: provides images and labels codes/curate_derm1m.py
Derm1M curation → DermLIP training: prepares training data codes/train_dermlip.py
DermLIP training → Research artifacts: writes model outputs README.md
Contagion gate → Research artifacts: writes gate features codes/stage1_gate.py
Gate evaluation → Research artifacts: reads cached features codes/stage1_report.py
Disease identification → Research artifacts: writes feature caches codes/stage2_disease_id.py
Caption de-leaking → Research artifacts: writes clean features codes/stage2_deleak.py
Trimodal evaluation → Research artifacts: reads clean features codes/stage2plus_trimodal.py
Zero-shot breadth → Research artifacts: writes evaluation results codes/zeroshot_breadth.py
Research queue → Research artifacts: writes queue results codes/run_queue.py
DermLIP training → ONNX export: supplies model components codes/export_onnx.py
ONNX export → Edge inference runtime: exports inference graphs README.md
Edge inference runtime → Triage and referral: routes triage outcome README.md

Drawn by GitDiagram on November 14, 2026 from the repository's file tree, README and source files. Read as Markdown ·