Pipeline complet Radiacode 103 - identification automatique d'isotopes

- VegaModel CNN-FCNN 34.5M params, 82 isotopes, val acc 99.89%
- Generation 50k spectres synthetiques 1D (12-24h durees)
- Entrainement 100 epochs sur RTX 5060 Ti (CUDA 12.8, Blackwell)
- Detection continue avec soustraction du background
- Capture background 24h avec gestion deconnexion
- Docker Compose : conteneur train (GPU) + detect (CPU/USB)
- Modele entraite inclus (vega_best.pt, 395 Mo)

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
Jacquin Antoine
2026-05-19 12:29:56 +02:00
commit 745a64b342
52 changed files with 17558 additions and 0 deletions

View File

@ -0,0 +1,82 @@
Ac-227
Ac-228
Ag-110m
Am-241
Au-198
Ba-133
Ba-137m
Be-7
Bi-210
Bi-211
Bi-212
Bi-214
C-14
Cd-109
Ce-139
Ce-144
Co-57
Co-58
Co-60
Cr-51
Cs-134
Cs-137
Eu-152
Eu-154
F-18
Fe-59
Ga-67
Hf-181
Hg-203
I-123
I-129
I-131
In-111
Ir-192
K-40
Lu-177
Mn-54
Na-22
Pa-231
Pa-234m
Pb-210
Pb-211
Pb-212
Pb-214
Po-210
Po-212
Po-214
Po-216
Po-218
Pr-144
Ra-223
Ra-224
Ra-226
Ra-228
Rh-106
Rn-220
Rn-222
Ru-106
Sb-125
Sc-46
Se-75
Sm-153
Sn-113
Sr-85
Sr-90
Ta-182
Tc-99m
Th-228
Th-230
Th-231
Th-232
Th-234
Tl-201
Tl-207
Tl-208
U-234
U-235
U-238
Xe-133
Y-88
Y-90
Zn-65